PUSH TRANSLATION 1

A REFERENCE MANUAL FOR THE ACADEMIC DRUG HUNTER

Version 1.0

Swami Subramaniam, MD, PhD

Member, Scientific Advisory Board, Ignite Life Science Foundation

Ramesh Jayaraman, MSc

Founder Director, DoseQuantics Consulting

1:  "Push translation" is a term used by us to signify the experimental efforts that academic “drug hunters” must undertake in order to prepare a newly discovered drug candidate for assessment by a commercial organization as a prospective licensing/investment opportunity

Feedback, comments, and suggestions are welcome.

Please write to drughuntersofindia@gmail.com

THE PURPOSE OF THIS MANUAL

Drug discovery is normally carried out in pharmaceutical companies. It is a multi-stage, multidisciplinary effort that takes many years and costs hundreds of millions of dollars. It is not possible to replicate the entire drug discovery-to-development process in academic settings. However, over the course of funding research proposals at Ignite Life Science Foundation, we come across projects where a candidate novel drug molecule with potential therapeutic applications is identified.

Examples of such instances include:

 (1) While exploring a biological mechanism whose modulation has therapeutic benefit, a chemical tool designed to probe the biology is identified as a potential drug

(2) A proposal is funded to improve the properties of an existing drug to overcome therapeutic failure—usually for diseases of national relevance, e.g., tuberculosis

(3) A known drug is repurposed for a new indication based on in silico structure-based screening

(4) Chemical analogs of a  known drug are synthesized in an effort to improve its properties

(5) A de novo drug design program to discover a novel drug for a new target is initiated, usually for a disease indication that is a national priority

Two government-driven initiatives have contributed to the increase in “drug hunting” efforts in academia: (1) an emphasis on translating the findings of basic research and (2) a special effort to address therapeutic needs unique to India, e.g., tuberculosis.

The end goal of academic research is publication. The requirements to publish experimental work are starkly different from the requirements of drug development. It is not the object of this document to turn academic investigators into professional drug developers. We expect most academic drug discoverers to license their discovery to a pharmaceutical company. Drug development is best left to the deep pockets and multidisciplinary skill sets available in the pharma industry. However, licensing an invention/drug candidate molecule to a biopharma company also requires some minimum work on the invention that reduces the risks associated with a novel molecule. Such a “developability assessment” of the molecule is required in addition to the core target/disease pharmacology of the drug. It includes answers to questions like "Is the drug manufacturable at a reasonable cost? Is the drug stable at room temperature? Is the drug soluble enough to be used by the preferred route of administration (critical for an intravenously administered drug)? etc. Some of these answers need additional experiments, although, given the development of in silico tools, many questions can be answered to a reasonable degree of approximation using these tools. Some of these tools are available free of charge, and we have included a table listing popular tools that are recommended by experts.

Since drug discovery is inherently risky, any reduction in perceived risk makes the drug more attractive in the eyes of a pharma company. Our objective in this manual is to offer guidance on a minimum set of experimental work/data that the investigators must generate in order to derisk their invention in the eyes of a potential licensee. We have taken preclinical proof of concept (pPoC) as the endpoint at which the IP asset can be passed on to a corporate entity for further development.

The required experiment/data to be generated at each stage is presented in the form of checklists at the end of each section. As the molecule progresses on its journey towards preclinical proof of concept (pPoC; this would correspond to TRL 3), the must-dos are those experiments that must be completed before progressing to the next stage. Bypassing these experiments will lead to avoidable grief in the form of uninterpretable data; for example, if PK data in the species used in efficacy testing is not obtained before doing the pPoC, the absence of efficacy in the pPoC could either be due to poor bioavailability or it could be due to the failure of the drug to produce a response despite engaging the target. There is no way to know which!

Apart from guidance on the experiments that are required, we also offer guidance on decision-making and prioritization during this phase of “push translation." For example, in competitive markets, where multiple drugs may be in development or the invention has to compete with already marketed drugs, a level of nuanced strategic positioning of how the new invention will improve on competing treatments, what significant and unique niches it will occupy, and how its advantages may be exploited for competitive benefit must be thought through. Why do we label this as “strategic”? Because the decisions are not guided purely based on scientific considerations. Prioritizing experimental work based on strategic considerations has to be done while keeping in mind commercial viability. This is a very different way of looking at things. While this cannot entirely supplant the need to do “good” science, we certainly see a role for such considerations in deciding what “must” be done versus what is “nice to do." For example, a once-a-day treatment as a competitive advantage over existing drugs that need to be taken three times a day may require a customized set of additional experiments, over and above proof of efficacy, that are specific to this strategic consideration.

A further consideration is the fact that scientists are often strapped for resources. In this situation, it is important that work be prioritized in terms of both money and time allocated. Time as an expendable resource is underemphasized in academia. But in drug development, time is a critical parameter. A delay of one month means one month less of sales, since the patent clock starts ticking from the date the patent application is filed. Delays can diminish the value of an asset significantly. Take a billion-dollar drug. A month of lost sales due to a delay in launch will be nearly 100 million dollars in lost sales, depreciating the value of the asset in the eyes of an investor.

SOME CAVEATS TO KEEP IN MIND

1. Every drug is different, and it is impossible to provide a generic list of considerations that apply to all. The guidance in this manual is broad and must be interpreted in the context of a specific drug being developed. Also, please bear in mind that the conflation of drug development science with drug development strategy means that the scientific rigor of the former has to be conditioned by the practical needs of drug development in an academic setting. The contents given here are the opinions of experienced drug developers and not textbook-style rules of drug development. Feel free to modify the experimental strategy guidance provided here, based on the specific circumstances that apply to your drug.
2. This manual mainly targets academic environments that are not equipped with all the bells and whistles needed for drug development. With respect to the needs specific to drug development, almost all academic laboratories can be considered low-resource environments. However, we do recognize that some laboratories and institutions may have greater access to drug development capabilities. When such capabilities are available, it would be prudent to make use of them and develop a fuller and more well-rounded package of experiments that is closer to what is expected in the pharmaceutical industry. The one consideration that has a bearing on doing this is speed. The patent clock is constantly ticking, and any long delays in doing the assessment and presenting to a potential licensee will depreciate the value of the asset.
3. We have worked back from the stage of completion of preclinical proof of concept (pPoC) to generate this guidance. Developing pPoC is a feasible objective for most academic laboratories (either within the same laboratory or through a collaboration) and is the minimum requirement to present a data package to a potential licensee. We have listed some CROs that can provide animal model services for a fee.
4. We have not provided detailed experimental method descriptions. Key references are provided.
5. Since this manual is only indicative and not definitive, we have also provided a list of domain experts who can be contacted for a phone consultation. The experts we have listed have indicated their willingness to provide advice/guidance pro bono, at least for an initial consultation.

ACKNOWLEDGEMENTS

A number of colleagues generously gave their time to read earlier drafts of this manual and provide comments. They are:

Rachit Agarwal, Associate Professor, Department of Bioengineering, Indian Institute of Science

Thomas Antony, Founder and Managing Director, BioPrompt Discovery Services 

Tanjore Balganesh, President, Gangagen Biotechnologies Pvt Ltd

Sowmya Bharath, Consultant Pathologist

Ranjan Chakrabarti, Executive Committee member, Federation of Asiatic Biotech Association; Advisory Board member, Pharma Now 

Javed Iqbal, Ex-Professor, IIT Kanpur; Ex-Director, Dr Reddy's Institute of Life Sciences; Ex-Head, Chemistry, Dr Reddy's Laboratories 

Harish Kumar MN, CSO, Leonid Chemicals Pvt Ltd

Radha Rangarajan, Director, Central Drug Research Institute

Ramesh Sistla, Founder and Managing Director at thinkMolecular Technologies Pvt. Ltd

Venkatesha Udupa, Senior Vice President, Toxicology, Glenmark Research Centre

Balasubramanian V, Co-founder & COO, Bugworks Research India Pvt. Ltd.

We are grateful to them for their help. Needless to say, we remain solely responsible for the contents, and any inaccuracies/errors are entirely our responsibility.

This book contains the chapters listed below. Feel free to click on hyperlinked text to go directly to the content of immediate interest. Some  chapters contain checklists as a convenient means to enumerate the critical things that need to be done. The checklist items in bold letters are the must-dos. These checklists are only intended as guidance. You may follow them based on your unique circumstances. But, if you choose not to, at least you are informed and forewarned. We have provided explanatory notes in the appendices to help understand the checklists. References appear as hyperlinked superscript numbers, and terms that need elaboration are hyperlinked to source material in the appendix or online. There is also a listing of the references at the end.

Chapters:

1. What makes a molecule a drug? (Checklist A)                        
2. The Target Product Profile (with an example)                        

3. Target validation(Checklist B)                                        

4. Completing the in vitro phase of drug discovery (Checklist C)                

5. Choosing an in vivo model (Checklist D)                                

6. Preparing for the in vivo experiment        (Checklist E)
7. Filing a patent: what, when, and how (Checklists F1 and F2)        
8. What an investor/licensee/collaborator looks for
9. How to find a licensee
10. A general checklist to qualify a drug for a second party (Checklist G)
11. Contingency planning (Checklist H)
12. Resources:

1. CROs
2.  Consultants
3. In Silico Tools

13. References
14. Appendix:

(1) The drug discovery process

(2) Explanation of commonly used terms in drug development

(3) Incorporating a company

(4)Why and where do drugs fail during development?

(5)Preclinical proof of concept (pPoC) is a key valuation inflection point

(6)What are Technology Readiness Levels?

WHAT MAKES A MOLECULE A DRUG?

Let us work backwards to answer this question: What are the distinguishing properties of clinically used and marketed drugs?

1. Efficacy in a disease or clinical condition—e.g., prevention or mitigation or cure
2. Relatively non-toxic (safe) in clinically used doses. Acceptable toxicity would vary depending on the condition being treated, e.g., cancer versus a sore throat, and hence, “RELATIVELY” non-toxic.

The two distinguishing properties stated above are quite obvious. But a few more conditions apply:

3. Feasibility—Let us say it will take 10 years of treatment or more to demonstrate clinical benefit; then no one will be interested in taking the risk of investing in expensive clinical trials on a drug that may or may not show benefit after a 10-year clinical trial. Such a drug is not developable unless there is a surrogate measure/biomarker that can be reliably used as an early predictor of efficacy. This was the case with statins, where lowering of LDL-C was used as a surrogate measure for cardiovascular morbidity.

4. Manufacturability—the drug ingredient can be made using methods that can be scaled up to address prevailing market demand AND the active ingredient can be formulated (i.e., for example, made into a tablet or injectable) using available technologies.

5. Supply Considerations—The drug must be resilient to the needs of the supply chain from the site of manufacture to the patient; e.g., it must be stable and it must be safe to transport

6. Clinical benefit/Value—The benefit/value the drug provides to the patient must be unique and desirable with respect to competing treatments.
7. Availability of Guidance Documents from Regulatory Agencies - USA (USFDA) and Europe (EMEA) have released, and continue to update guidance documents recommending the preclinical and clinical studies that are needed for drugs to be approved. The Indian drug regulator (CDSCO) also has some broad guidelines, but nothing specific to new small molecule drug development. Such a consideration is especially important for a first-in-class drug, for which there is no precedent. For example, if the plan is to use a surrogate endpoint (a blood biomarker) to establish efficacy, then the use of the biomarker in such a fashion must be pre-approved by the regulatory body (this applies to a drug in clinical trials and is not a hard requirement for pPoC studies). There are situations where the field is nascent and guidance documents have not been published. In such a case, approval for a development plan can be sought in discussion with the regulatory authority prior to the start of clinical development. There must be good reasons to believe, based on the science, that such approvals will not be withheld.
8.  Adequate ROI—There must be a significant return on investments (ROI) made in R&D, manufacturing, and marketing. To enjoy such an ROI, the marketing company must have exclusivity via patent rights. In addition, there must be adequate pricing elasticity to obtain a targeted ROI. This, for example, is not the case for extremely expensive medicines like gene therapy, which are priced at a level (millions of dollars per treatment)  where there is nil elasticity due to the high cost of development that has to be recovered from limited units saleable versus affordability. Additionally, gene therapies for monogenic disorders are single-shot treatments, and post-successful gene therapy, there is no additional consumption of the treatment. This means the market is relatively small and demand elasticity is absent. In the absence of demand elasticity, there is no scope to increase sales volumes by decreasing prices. Similar, although less punishing, ROI considerations apply to antibiotics, where a few doses can produce a complete cure and the newest antibiotics are often moved to the reserve list (i.e., they cannot be prescribed unless older antibiotics have been used and found ineffective), thus limiting market size, apart from the ever-present danger of antibiotic resistance, which can make a much-vaunted antibiotic useless. It is prudent to assess the ROI for the envisaged invention using the above criteria before committing resources.

Now for the CHECKLIST. Here is one screen you can apply to your early-stage discovery that can provide sufficient justification for doing further work or even showcasing the discovery to a potential licensee.

CHECKLIST A: Is your molecule “potentially" a drug (experimental methods to answer some of these questions are given in the following sections)?

• Is it effective in a predictive animal model of human disease at doses/concentrations likely achievable in humans?
• At effective doses/concentrations in the disease model, is there data/evidence to show that it is likely to be safe?
• Is there a clear and feasible pathway for development/regulatory approval?
• Can the invention be protected by patents?
• Is there a sufficiently large marketing opportunity (i.e., ROI) to justify development costs?

A standard approach followed in industry to link preclinical experimental work to the needs of eventual clinical approval is to develop what is called a "Target Product Profile" (TPP). TPPs can be very comprehensive, since the items cover the gamut of steps in discovery and development all the way to clinical approval. In the next section, we provide an example of a shortened version of the TPP that governs requirements that must be fulfilled up to preclinical proof of concept. You should create your own TPP based on the requirements you are targeting with your molecule.

THE TARGET PRODUCT PROFILE

A Target Product Profile (TPP) in early discovery acts as a strategic "North Star," outlining mandatory and ideal product characteristics (safety, efficacy, dosing, population) to guide development, align with regulatory goals, and attract investors. It is a structured document that defines the end goal to optimize candidate selection and reduce development risks early. A TPP is usually drawn up at the beginning of the discovery program to serve as a guide for the preclinical stage to be in alignment with the clinical requirement. The TPP is a living document that is continuously updated through the project lifecycle in response to emerging data from competition and target biology.

The TPP for an early-stage academic discovery project, as being considered in this manual, might be less specific on things like clinical dosing and regulatory goals unless they are the key competitive advantages for positioning the molecule in the eyes of a buyer. The TPP should contain the key elements that an investor might look for. An example checklist of items is as follows:

1. Evidence of efficacy in an animal model that is predictive of clinical disease X along with targeted route of administration, targeted Emax and targeted EC50, e.g., oral administration with doses in the range of 3-30 mg/kg once a day and reduction of blood sugar, either to control levels or to 70% of pretreatment fasting levels
2. Evidence of superiority over existing/competing (standard of care) treatments, e.g., produces complete resolution of disease pathology in 3 months versus 8 months for the competing product
3. Safety in 2-3 models (in vitro plus in vivo)

A key nuance that will determine the items in the TPP is whether

1. The molecule is claimed to be BEST IN CLASS

A "best-in-class" (BIC) drug is a follow-on drug that demonstrates superior efficacy, safety, or convenience compared to existing, approved drugs that act on the same target as previously marketed drugs (or drugs in late-stage development on which there is sufficient publicly available information so as to make an informed comparison)

Key Characteristics of Best-in-Class Drugs:

• Superior Efficacy/Safety: Provides better clinical outcomes or fewer side effects than drugs in the market or in late-stage development

AND/OR

• Improved Compliance: Enhanced, easier dosing schedules (e.g., once-a-day vs. three-times-a-day)

AND/OR

• Target Optimization: Often optimized for better binding potency or pharmacokinetics compared to the original drug

2. The molecule is claimed to be FIRST IN CLASS

A first-in-class (FIC) drug is one that uses a new and unique mechanism of action to treat a medical condition, differing from existing therapies by targeting previously untargeted molecules. These drugs act on a new molecular target or use a novel pathway to affect a disease, rather than just improving upon existing drug classes.

Now let us look at a sample TPP. You must construct your own TPP based on the properties you wish to target for your molecule.

Example of a Target Product Profile for a drug for acute hemorrhagic stroke

A target product profile for a Best in Class molecule should mention the property and benchmark performance on the basis of which this molecule is better than existing drugs of the same class. For example, if the claim for the new molecule as Best in Class is made on the basis of superior efficacy, then item 4 in the table above could read as follows: “an improvement of at least 30% in area protected compared to the existing drug.” If the claim that is made is that the new molecule is Best in Class because it has some desirable properties, such as a longer plasma half-life, then evidence needs to be generated supporting this claim using human liver microsomes to show that metabolism is considerably slower, potentially leading to a longer t1/2 in humans with the new drug, permitting less frequent dosing (at this stage direct clinical data with this drug cannot be generated since it has not passed the stage of regulatory approval for human studies).

For a First in Class molecule, it should be clear from the TPP what the advantage of this molecule is over the current standard of treatment, including drugs that act through other mechanisms. Since "First in Class" implies an unprecedented target for the molecule, it should be clear in the TPP what advantage this unprecedented mechanism of action provides. For example, for a novel antibiotic, the new mechanism of action may make the drug effective against organisms that are resistant to all previously known antibiotics.

TARGET VALIDATION

Target validation is a critical drug discovery step that confirms that a specific molecular target (usually a protein receptor to which the drug binds) is directly involved in the disease process being targeted and that modulating the target will have a desirable therapeutic effect. A validated target helps anticipate the efficacy and safety profile of a novel drug, and such a drug is less risky to develop compared to a drug whose target is either completely unknown or incompletely understood (unvalidated). Most discoveries made in academic laboratories will be associated with knowledge of the biological target, since target biology is often the starting point for academic projects. However, knowing the target does not mean that the target is validated for a particular clinical application. The ways in which a target is validated are enumerated in checklist B (see below).

What if I do not know what the molecular target of my drug is? This could happen when the drug has been identified through a phenotypic assay, e.g., it modulates the behavior of a cell, organ, or organism suggestive of a beneficial effect on disease. In the current environment for drug development, it will be nearly impossible to attract an investor given the high risks associated with a drug with an unknown molecular target. An exception may be a repurposed drug. Let us say you find that the well-known antidiabetic metformin, which has a long history of clinical use, reverses aging in mice. You could file a method-of-use patent for such a new indication for metformin. Such method-of-use patents are weaker and less preferred compared to composition-of-matter patents. On the other hand, if we knew that metformin was producing its effect through a validated target, it would be useful information towards designing analogs of metformin optimized for the new clinical application. Such analoging can yield patentable molecules, a necessity if an investor must show interest. See here for some standard methods to identify the target for a drug.

CHECKLIST B: When can I consider a target validated?

• Is there evidence of target involvement in the disease?, i.e., the association of the molecular target with the disease—e.g., is it present in the diseased tissue and organ, and does changing its expression (knockout or knockdown) influence disease processes in the expected fashion? Are changes in activity/expression of the molecular target associated with the clinical disease state? E.g., if a mutation is causing disease, is the mutation found in the target protein’s gene? An example is the association of the BCR-ABL fusion kinase, a genetic variant, with the development of chronic myelogenous leukemia, against which the drug Gleevec was developed.
• Is the target druggable? Does it have a 3D structure that provides opportunities to design small molecules that can bind and interact with it with high affinity (pM to nM)? Will drugs that modulate the activity of this target have a sufficient margin of safety and acceptable toxicity?
• Has it been shown that pharmacological modulation of the target has an effect on the disease process and that such effects can be measured using available biological tools/assays? Do such drugs also show beneficial effects in animal models of the disease that are predictive of clinical utility in humans? Is there a clear dose-response relationship between the drug and target activity?
• When the drug modulates target function, is there sufficient margin of safety between desirable modulation and modulation that leads to unacceptable toxicity?

COMPLETING THE IN VITRO PHASE OF DRUG DISCOVERY

1. In vitro drug discovery: This stage involves studies done to characterize the candidate drug molecule’s pharmacology (in vitro), toxicity, and drug-likeness 1 . This part is usually well within the scope of the capabilities in academia, either within the inventor's lab or via a collaboration.
2. Target engagement - CETSA (cellular thermal shift assay) can be used as a means to detect target engagement and should be accessible to any lab using classical technical methods. If an endogenous ligand for the protein target exists, then it is important to know if the drug competes with the endogenous ligand for the same binding site OR if the drug modulates the protein at a different site (allosteric).
3. It is very useful to know the site on the target protein where the drug binds. With the help of AI tools like AlphaFold, structures of proteins can be computed with reasonable accuracy with an option to do a molecular dynamics simulation to identify putative sites where the molecule interacts with the target protein. This information is valuable from the following standpoints: (a) if the identified interaction is a canonical binding site on the target for previously studied drugs, then it gives comfort with respect to efficacy and the potential for on-target toxicity. (b) the binding site interaction will also provide information on the “pharmacophore” on the drug molecule that must be kept in mind if any further modifications are proposed for improving physicochemical properties
4. Access to target compartment—If the target is sequestered within a hard-to-access compartment (e.g., brain, eye, or inside a cell), then it is important to know that sufficient concentrations of the molecule are reached in the compartment where the target is present. Initial data from cell-free assays has to be recapitulated in cell-based assays to determine access across the cell membrane. An estimate  of protein binding, 7, 8 will help determine the amount of free drug available to bind to the target. Please see this video primer on the importance of measuring free drug: Not a Hypothesis: The Free Drug Principle
5. Estimating efficacy and potency: A full concentration-response curve (8 concentrations of the molecule spaced logarithmically, e.g., 0.1, 0.3, 1, 3, 10, 30, 100, 300 µM) is needed to calculate the efficacy (Emax) and potency (EC50 or IC50) 2 .
6. Time effects: The pharmacological effect of a drug molecule can be instantaneous 2 or delayed over time depending on the pharmacological effect following drug-target engagement 3 . Time-dependent effects should be measured (in cell-based systems). Biological effects may cumulate over time, a likely scenario where the in vivo exposure may be prolonged due to the maintenance of blood levels for longer periods than those studied in in vitro assays or because of the effect persisting even after the drug is completely eliminated due to long-standing changes at the target level (for example, by a drug that binds covalently to the target). 
7. Reversibility—It is important to know whether the binding of the drug to the target and/or its biological effect is fully reversible. Drugs that bind covalently (and are therefore irreversible) would be problematic for chronic treatment of conditions like hypertension. Covalent binding may be acceptable for some specific classes of drugs.
8. Lack of toxicity to human cells or organoids—choose a human cell line or primary cells. Toxicity should not be observed at up to 10-50 times the EC50
9. Need for a bioanalytical method—Many of the above assays will call for a bioanalytical assay method that can be used to accurately measure drug concentrations in a variety of biological matrices, starting with cell extracts and cell culture medium. The method should be able to measure drug concentrations 100X below the EC50, be specific, show a linear response at the detector, and ideally be fast and cheap.
10. Is the molecule sufficiently soluble for a solution to be prepared for administration by the preferred route in animal model 4 so as to achieve the targeted dose/blood concentration? If DMSO is used as a solubilizer, then the final concentration of DMSO in the formulation to be administered should not exceed 0.5%.
11. Stability of molecule : Is the molecule stable in plasma and in liver microsomes prepared from the species being used for the pPoC? High instability in plasma and in liver microsomes is predictive of rapid elimination of the molecule in vivo 5 . Is the molecule stable for the duration of the in vivo experiment in the formulation used for dosing (if stock solutions are to be prepared and stored, then it should be checked that the molecule is stable throughout the process until it is administered)?
12. Analoging the lead compound to develop improved/patentable variants as a way to de-risk the possibility that the lead compound fails. In silico SBDD (structure-based drug design) can help do this. Medicinal chemists recognize from experience that certain structural functionalities are not desirable in a drug. If there is a series of compounds that have been made, then there is an option to weed out such compounds from the series.

CHECKLIST C: What is the minimum data required before embarking on in vivo testing in an animal model?

• Has the molecular target been identified/confirmed?
• Has the structure of the drug been robustly characterized?
• Is the drug substance used in the assay of sufficient purity to relate effect to the drug (aim for purity in excess of 95%)
• Is there evidence that the molecule has access to the target in the in vitro assay? - e.g., if the target is intracellular, does the molecule cross the cell membrane and achieve adequate intracellular concentration (check if it is a substrate for efflux pumps that operate in cells)
• Is there a good fix on the in vitro EC50 and Emax? I.e., are the estimates of these parameters sufficiently precise to allow PK-PD correlation in the pPoC?
• Is the effect reversible? - again, this is important while correlating PD with PK
• Is the molecule drug-like?
• What is the duration of exposure required for optimum pharmacological efficacy?
• Is the effect sufficiently specific? Has the molecule been evaluated in a counterscreen?
• Is it non-toxic to human cell lines at concentrations at least 10-50X the EC50
• Is there a bioanalytical method capable of reproducibly measuring concentrations between 1/100th and 100 times the concentrations likely to be encountered in samples from experiments?
• Pharmacokinetics: Is the concentration attainable and duration of exposure in the compartment where the target resides sufficient to generate the desired response? This can be determined in a single-dose PK study in the animal species to be used for the pPoC by the route of administration chosen for the in vivo study. Ideally, the sampling times should be a total of 8-10 after the dosing (e.g., for an orally administered drug: 5, 15, 30, 60, 120, 240 min, and 24 h and 48 h post-dosing; a sample protocol can be found here). The peak concentration in blood/tissue (Cmax), the time of peak concentration (tmax), the area under the plasma concentration time (AUC), and the elimination half-life (t1/2) generated from this study will help confirm that adequate free concentrations (e.g., fCmax/EC50, fAUC/EC50 or time above EC50) of the drug are achieved in the biological compartment in which the targeted receptor is present6. In conjunction with information on protein binding 7,  8 an estimate of free drug available to interact with the target can be made.

CHOOSING AN IN VIVO MODEL

Preclinical proof of concept (pPoC) in a disease model

Preclinical proof of concept (pPoC) in a disease model is a critical phase in drug development that provides evidence that a new therapeutic candidate works as intended—demonstrating efficacy and safety in non-human models (animals or human-derived systems) before testing in humans. It bridges the gap between basic discovery and clinical trials by answering whether the treatment has potential or should be abandoned, thereby reducing the risk of costly late-stage failures.

There are many non-animal models that are emerging, and the USFDA actively encourages the use of these models (e.g., 3D organoids, in lieu of animal models) for evaluating  efficacy. However, the deployment and use of such models has to be limited to those that have been adequately validated in the published literature. Even when not validated, human organoid models can be used to complement data obtained in more traditional animal models10. Most of the comments that follow can also be applied to the use of such models.

Key Considerations for a Successful pPoC

1. Model Selection & Validation: The model must reproduce key aspects of human pathophysiology and pharmacology. It must contain the molecular target for the drug being tested in a context similar to the context in humans (e.g. if the drug is an antihypertensive intended to act on receptors in vascular smooth muscle in humans, then the same receptor must be present in the vascular smooth muscle in the animal model as well).
2. Experimental Design: Studies should be blinded, have appropriate positive/negative controls, and be repeated to ensure reproducibility.
3. Dose-Exposure-Response Relationships: The study should define the effective dose and exposure range, including the minimally effective and optimal biological (exposures) doses.
4. PK/PD relationships: Integrated pharmacokinetics (PK) and pharmacodynamics (PD) insights are essential to relate drug exposure to biological response. This provides a quantitative basis for fixing dose, regimen, and duration.
5. Timing of Administration: Evaluation of the drug's effect relative to the onset of disease or injury. PK/PD time course information will help understand the relationship between drug concentration and onset, intensity, and duration of effect and, consequently, form the basis for administration timings.

Key Components of Preclinical PoC

A successful pPoC study typically establishes three main outcomes:

• Efficacy: That the treatment demonstrates the desired biological (pharmacological) effect, such as reducing disease symptoms that can be associated with its PK in a dose-dependent manner (quantitative PK/PD relationship), in an in vivo model.
• Mechanism of Action: The drug interacts with biological systems and pathways in the intended manner.
• Safety/Tolerability: The treatment should produce minimal adverse effects at therapeutic doses, e.g. A therapeutic index of 2+ or minimal clinically observed adverse effects at therapeutic doses. The margins of safety proposed here will vary depending on the clinical use of the drug. A drug for cancer may be acceptable with a lower margin of safety compared to a drug for hypertension or dyslipidemia.

Advantages and Regulatory Context

• Risk Reduction: Identifies flawed candidates early, avoiding high expenses of failed human trials.
• Attracts Funding: Provides hard data necessary to secure investment.
• Informs Clinical Trial Design: Helps determine initial patient populations, doses, and routes of administration for Phase I trials.
• Regulatory Support: Although a GLP study is not a requirement for the purpose served by this manual, data from well-conducted, GLP (Good Laboratory Practice) pPoC studies are crucial for Investigational New Drug (IND) applications.

CHECKLIST D: When can I consider a preclinical model sufficiently predictive of efficacy in human clinical disease?

• Does the model contain the molecular pathway/target that has to be activated or inhibited for drug efficacy in the same organs, and do the molecular pathways behave in response to drugs in a manner that supports predictive efficacy in the human disease?
• Does the model reasonably replicate the phenotype of the human disease?
• Do drugs shown to be effective for the disease show activity in the model?
• Is the rank order of efficacy of a series of known drugs in the animal model mirroring the rank order of efficacy in humans?

PREPARING FOR THE IN VIVO EXPERIMENT

Demonstration of preclinical Proof of Concept (pPoC), i.e., translation of in vitro potency of the drug to in vivo efficacy (biological/pharmacological effect) in the animal model of disease, is a critical stage of the drug discovery project. Failure to demonstrate pPoC can cause termination of the program. Therefore, it is very important that the in vivo efficacy study be carefully designed based on the available in vitro, pharmacokinetic, and safety data for the candidate compound. The following checklist should be helpful in experimental design. Ideally, pPoC should be generated in more than one model, preferably in two different species.

1.         Doses for the in vivo study: The selection of doses should be based on in vitro potency (IC50 or EC50), efficacy (Emax), pharmacokinetics (PK), and safety. Ideally, one should select at least 4 to 6 dose levels (a minimum of 3 dose levels) for understanding dose-response relationships. The doses selected should be safe, display a linear increase in exposure (Cmax, AUC), and at the minimum achieve pharmacologically active exposures. Pharmacologically active exposures are concentrations that demonstrate efficacy in in vitro cell-based assays expressing the pharmacological target. As the pharmacological effect of the compound will be measured under varying concentrations in vivo (blood, serum, plasma, and tissue), in contrast to the in vitro experiments where the effect is measured against a constant concentration, it is important to ensure that the in vivo concentrations exceed in vitro potency at a level and for a duration sufficient to engage the target to produce the desired beneficial effect. A benchmark to aim for would be a blood concentration that is 3-fold higher than the EC50 observed in in vitro experiments and maintained long enough for a desired effect duration. The latter would depend on the PK of the drug, particularly the speed with which it is cleared from plasma through metabolism, renal excretion, etc. The targeted concentration and duration will also affect the probability of seeing effects that presage toxicity. The ratio of the highest dose or exposure (plasma concentration, AUC) that does not produce an adverse effect (No Observed Adverse Effect Level; NOAEL) over the exposure (plasma concentration, AUC) required to produce the desired pharmacological effect would be the therapeutic index. A high therapeutic index would reflect a high safety margin (preferably 10-50x). If the therapeutic index is equal to or less than 1, the drug is not developable, with the exception being drugs for certain incurable (and fatal) conditions like cancer, where a monitorable toxicity biomarker can be used to titrate doses such that benefit exceeds risk. In certain instances the observation may not hold true in humans, and hence the drug may still be developed.

PK data should be available before the start of the pPoC. If PK data is not available for the selected doses, measuring PK in the pPoC study can help interpret the efficacy results using exposure data. If a negative result is obtained, it can be determined if it is a failure of the mechanism or the absence of target engagement.

2.         Compound amounts: Ensure compound amounts are sufficient to be used for the in vivo experiment. In vivo experiments can be short in duration (24 h) or moderately long (a week) or long (1 month or more). The compound amount will depend on the dose level, number of dose levels, dosing frequency, and duration of dosing. It is critical that the amount of compound  be sufficient to meet these requirements.

3.         Formulation: Select the appropriate vehicle for the preparation of the formulation of the compound for dosing. The vehicle should be safe for the intended route. Refer to best practices in literature for selection of the appropriate vehicles based on the physico-chemical properties of the compound. If it is parenteral, the compound should be soluble in the vehicle for the intended doses. If not soluble, the intended doses cannot be administered as planned, or the compound can precipitate following administration, leading to sub-therapeutic exposures. If it is oral,  ideally soluble formulations are preferred, but if not possible, suspensions are also suitable based on PK.  

4.         Validated model: The in vivo pharmacological model should have been validated with proper controls and reference drugs and pharmacodynamic endpoints. Ensure availability of adequate amounts of reference drugs or reference compounds for the study. If the model is not available in-house, it can be outsourced to a CRO that has the validated model.

5.         Experimental design: Last but not the least. The experiment should be designed appropriately to draw meaningful conclusions. A well-designed experiment can save multiple experiments, repetitions, or equivocal results. Appropriate controls such as vehicle, reference, and treatment groups with adequate numbers of animals per group for statistical power should be included.

6.         Testing in more than one in vivo model (ideally in different species) is highly desirable at this stage

Checklist E:  Considerations for testing in the preclinical pPOC model

• Is the drug formulation stable in the conditions and for the duration of the experiment?
• Do the chosen dose(s) for pPoC to show efficacy produce drug concentrations in the target compartment in the range of the 3-5X the EC50 or IC50
•  Is a formulation available that can deliver the required dose by the chosen route? E.g., for an IV formulation, the drug should not precipitate in the blood and there should be no or minimal local toxicity at the site of administration
• Is the drug stable in the biological matrix of the target organ—plasma, urine, or brain tissue, depending on where the drug is supposed to act?
• Is single-dose/multiple-dose PK data available for a minimum of 3 dose levels (say, 1, 3, and 10 mg/kg doses selected bracketing the dose most likely to show optimum efficacy?
• Is an adequate amount of test compound available based on the experimental design (including necessary replicates)
• Is the experimental design adequate for meaningful interpretation of results

FILING A PATENT: WHAT, WHEN, AND HOW

Patenting (owning intellectual property rights (IPR)) a molecule (NCE or NBE) or a class of molecules is mandatory for a) protecting the company’s assets and therefore preventing others from working on the same molecule or class and b) commercialization, either by out-licensing or by revenues following marketing approval. The lack of IPR does not attract investments and negatively impacts out-licensing opportunities. Therefore, it is critical to patent promising molecular assets at the appropriate time. The following considerations must be kept in mind:

1.         Novelty: Ensure that the drug is truly novel. The medicinal chemist of the project should ensure that the molecules designed are novel in structure and do not infringe on patents of others based on an extensive literature and patent survey. This can be assisted with expert inputs from the institution’s patent cell (CSIR labs can use URDIP). It would be helpful to get expert help (if not available internally, then through patent attorneys) that has experience in pharmaceutical sciences for patent search and filing. An experienced patent attorney can ensure the novelty of the assets by doing the appropriate searches and supporting technical arguments.

2.         When to patent: Ideally, the patent should be filed following identification of the lead molecule and its analogs (the lead series). Public disclosure of the structure of the drug in a patent or a presentation will invalidate a subsequent patent application. The need to claim priority before being scooped by the competition is also a determining factor in timing the filing.

3.         What to patent: It is best to file for a product (composition of matter) patent, as it has commercial attraction. File for a patent only when the molecule or molecular class a) is novel in chemical structure; b) displays in vitro activity on pharmacological targets outside and inside cells, and c) has drug-like properties (in vitro ADME and safety and PK). Patents for use can also be filed for existing molecules, provided that the use has been adequately and sufficiently differentiated for a therapeutic use that was not obvious previously.

4.         Data integrity for patenting: Ensure that all the experiments performed with the NCEs are properly recorded, signed, and dated and witnessed by the investigators in the laboratory notebooks. The books must be linked to raw data for the appropriate studies. The laboratory notebooks and raw data should be corroborated by a member who is not involved in the experiments. The claim for priority of invention will be based on the entries in the lab notebooks (digital notebooks can be used). If not properly recorded, the claims for novelty can be dismissed or become invalid in the court.

Here is a video primer on patenting small molecules: Patenting Strategies for Small Molecule Drugs

CHECKLIST F1: Requirements for filing a provisional patent application

• Has the novelty of chemical structure been ascertained?
• Do the specifications listed in the patent  account for all planned improvements, e.g., wider application of the invention, backup molecules,
• Has a patent attorney been identified for filing a patent?
• Is the data for patenting complete?  A Lead molecule with desirable biological property has been identified and its structure has been verified using multiple means
• Have all data been appropriately recorded in a signed and dated notebook and stored for filing

CHECKLIST F2: Requirements for filing a final patent application (in addition to requirements mentioned for filing a provisional patent application)

• Has the novelty of chemical structure been thoroughly established using professional help to do a Freedom to Operate (FTO) search?
• If the chemical compound is a racemic mixture, has the patent claim specified which racemic form is active?
• Has a patent attorney been identified for filing a patent with global coverage?
• Is data for patenting complete for the identified lead?

WHAT AN INVESTOR/LICENSEE/COLLABORATOR/FUNDING AGENCY MIGHT LOOK FOR

There are two potential endgame scenarios for an academic inventor—

1. In the first instance the academic inventor may want to try developing the asset a little further. Capital could be raised for this work by incorporating a biotech company and vesting the IP in that company. The IP could be developed by this company to a point where attractive terms can be negotiated with a major pharma to outlicense the drug. Practically speaking, such an approach is hard to execute for an academic faculty for two reasons: inexperience with the various processes needed to execute such a plan (e.g., capital raise), the inability to commit the time needed, and the inability to leave the faculty position to become part of the new company due to the risk of failure. A situation where such an approach might work would be if the invention were a platform technology, e.g., a drug delivery system. In such a case, while the platform remains within the new company formed, it can be used to strike multiple deals with pharma companies for products they are developing. The counterparties to such deals would assume most of the risk of developing the combination of the platform technology with their drug.
2.  Alternatively, the investigator could find an investor early on who will assume the risk and cost of further developing the IP to sell to a pharma company, and this is usually a biotech working on products for the same clinical indication or adjacent clinical indications (major pharma companies are usually uninterested in an early-stage asset with no clinical data).

It is important to position the invention attractively by demonstrating its market acceptability. A first-in-class drug for an unmet clinical need (especially one that has a fatal outcome, e.g., pancreatic cancer) would be an example of high market acceptability. If the drug is not first-in-class, then a best-in-class positioning can also be attractive for an investor. For example, if the drug has a clinically significant improvement in efficacy, e.g., 90% of patients respond satisfactorily versus only 75% with the best standard of care treatment. Lesser advantages that can still be desirable enough to quickly make a follow-on drug a leader in the category are things like once-daily dosing versus multiple daily doses for the competitor or fewer side effects. A cheaper drug substance may not be a key competitive advantage since for most drugs under patent the extra cost of manufacturing can easily be subsumed within the price. Early consideration of positioning advantage will ensure that the critical experiment to robustly prove this advantage is a part of the package supplied to a potential licensee.

HOW TO FIND A LICENSEE

Most large universities/research institutes run a tech transfer office that is accountable for finding licensees. However, since they may lack specialized expertise in the domain of the invention, they will often need the support of the inventor to find and establish connections with potential licensees. Since inventions originating from academic laboratories have limited data up to pPoC, large pharmaceutical companies may not be interested. The most likely investors would be companies from emerging markets (including China) that seek to bolster their limited internal efforts or small- to medium-sized biotech companies seeking to expand their discovery pipelines. The best way to find them is to register for one of the Biopartnering meetings that take place in Europe (e.g., Bio-Europe: https://informaconnect.com/bioeurope/) and the US (https://bpjw.bio.org/). Retaining an experienced consultant with past experience in biopharma licensing and dealmaking will increase the probability of success and help obtain better value for the asset. The Government of India also has schemes to support early translational research, and these sources can be explored before licensing to an external party. Any external party will seek to repeat key tests in their own laboratories or via CROs under a material transfer agreement (MTA).

CHECKLIST G: A GENERAL CHECKLIST TO QUALIFY A DRUG IN THE EYE OF AN INVESTOR (the items in this checklist are desirable but may not always be attainable in the academic environment; the items in bold are a must-have)

• Proof of Mechanism of Action/Target Engagement
• Proof of efficacy in a predictive animal model with a dose response relationship
• Proof that the drug is safe/non-toxic in the clinically useful dose/concentration (minimally against a human cell line plus a therapeutic index of 2+ in the pPoC animal model)
• Proof of superiority versus standard treatments (if a standard treatment exists)
• Proof of bioavailability that is clinically useful
• PK-PD correlation: Efficacy at concentrations achievable in humans
• Physicochemical properties compatible with requirements for commercially viable manufacture, formulation and long-term stability
• A defined regulatory path for drug approval
• Market attractiveness
• IP protection for the invention

The presentation to the licensee should contain the sections mentioned in the checklist above, and a standard pitch deck format can be found here.

CONTINGENCY PLANNING: How to prepare for failure or changes that may influence outcomes and objectives

A drug discovery program may be seriously affected when one or more of the following situations occur: a) failure to demonstrate pPoC; b) unexpected toxicity; c) the candidate does not show non-inferiority or superiority to the competitor; d) a competitor has filed a patent that is similar to the candidate compound.

While the failure to demonstrate pPoC is a setback, it is worthwhile to analyze and better understand the pharmacokinetic/pharmacodynamic (PK/PD) reasons for the failure. Possible reasons could be sub-optimal dosing resulting in a lack of adequate exposure at the site of pharmacological action or a delayed effect (delayed onset of effect following dosing). Understanding the reasons for failure can be of critical help to design the next series of compounds. Unexpected toxicity can be analyzed by studying potential off-target effects using  in vitro or cellular assays that were not studied before or by obtaining information on chemical classes similar to that of the candidate.

To minimize the adverse impact, the following measures can be considered or followed : a) The possibility of performing a second efficacy experiment with scope for improved experimental design should be planned in advance as a contingency measure; b) It is prudent to have backup compounds to the candidate that have at least one property that is, if not better, then at least similar to the candidate drug; c) In the case of patent infringement with a competitor, it is best to have medicinal chemistry plans in place to achieve novelty. This may delay the project but help keep it alive.

CHECKLIST H: Requirements for contingency planning

• Are compound quantities and related resources adequate for repeat studies or for testing in an alternative model?
• Have backup compound(s) been identified from the same or different chemical series?
• Are medicinal chemistry plans in place for overcoming patent infringement issues that may disqualify the lead candidate?

RESOURCES

Drug discovery is a multidisciplinary activity involving scientists from different disciplines such as biology, biochemistry, molecular and cell biology, computational science, medicinal chemistry, bioanalytics, pharmacokinetics, pharmacodynamics, pharmacology, and toxicology. Given the nature of academic research, it is unrealistic to expect professionals from all the above disciplines to be available with access to the appropriate instruments and facilities. Thus, academic drug discovery groups may need to collaborate with professionals and laboratories with expertise.

Useful resources for collaborations are consultants in different fields of drug discovery and development; contract research organizations (CROs), which offer services in drug discovery; central laboratories, which offer all services under one roof; or other academic laboratories.

CROs

INDEPENDENT CONSULTANTS IN DRUG DISCOVERY (Link to linkedin profiles)

The following consultants have agreed to provide at least one hour of their time on a pro bono basis for a preliminary consultation. If you have direct drug discovery experience and would like to volunteer to be included in this listing, please do write in.

IN SILICO TOOLS

REFERENCES

1. Smith DA, Jones BC, Walker DK. 1996. Design of drugs involving the concepts and theories of drug metabolism and pharmacokinetics. Med Res Rev. May;16(3):243-66.
2. Holford N. 2017. Pharmacodynamic principles and the time course of immediate drug effects. Transl Clin Pharmacol. Dec;25(4):157-161.
3. Johan Gabrielsson, Hugues Dolgos, Per-Göran Gillberg, Ulf Bredberg, Bert Benthem, Göran Duker. 2009. Early integration of pharmacokinetic and dynamic reasoning is essential for optimal development of lead compounds: strategic considerations. Drug Discov Today. 4(7-8):358-72.
4. Neervannan S. 2006. Preclinical formulations for discovery and toxicology: physicochemical challenges. Expert Opin Drug Metab Toxicol. Oct;2(5):715-31.
5. Kumar GN, Surapaneni S. 2001. Role of drug metabolism in drug discovery and development. Med Res Rev. Sep;21(5):397-411.

6. Jang GR, Harris RZ, Lau DT. 2001. Pharmacokinetics and its role in small molecule drug discovery research. Med Res Rev. Sep;21(5):382-96.
7. Pacifici GM, Viani A. 1992. Methods of determining plasma and tissue binding of drugs. Clin Pharmacokinet 23:449–468.
8. Waters NJ, Jones R, Williams G, Sohal B. 2008. Validation of a rapid equilibrium dialysis approach for the measurement of plasma protein binding. J Pharm Sci. 97(10):4586-95
9. M3(R2) Nonclinical Safety Studies for the Conduct of Human Clinical Trials and Marketing Authorization for Pharmaceuticals. Guidance for Industry. Department of Health and Human Services Food and Drug Administration, Center for Drug Evaluation and Research (CDER), Center for Biologics Evaluation and Research (CBER). January 2010 . Revision 1 ICH
10. Indiwari Gopallawa, Charu Gupta, Rayan Jawa, Arya Cyril, Vibha Jawa, Narendra Chirmule, Vikramsingh Gujar, Applications of Organoids in Advancing Drug Discovery and Development, Journal of Pharmaceutical Sciences, Volume 113, Issue 9, 2024, Pages 2659-2667, https://doi.org/10.1016/j.xphs.2024.06.016.
11. A terrific and growing collection of video primers that discuss concepts relating to drug discovery and development can be found here: https://www.youtube.com/@drughunter/videos.
12. This reference targeted at academics and startups (US-centric) lists milestones, timelines, and costs at each stage of discovery and development for cancer drugs: Trovel J, Sittampalam S, Coussens NP, et al. Early Drug Discovery and Development Guidelines: For Academic Researchers, Collaborators, and Start-up Companies. 2012 May 1 [Updated 2016 Jul 1]. In: Markossian S, Grossman A, Baskir H, et al., editors. Assay Guidance Manual [Internet]. Bethesda (MD): Eli Lilly & Company and the National Center for Advancing Translational Sciences; 2004-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK92015/ 
13. Drug Discovery in an Academic Setting: Playing to the Strengths,  Donna M. Huryn
    ACS Medicinal Chemistry Letters 2013 4 (3), 313-315, DOI: 10.1021/ml400012g

14. Murray AJ, Cox LR, Adcock HV, Roberts RA. Academic drug discovery: Challenges and opportunities. Drug Discov Today. 2024 Apr;29(4):103918. doi: 10.1016/j.drudis.2024.103918. Epub 2024 Feb 14. PMID: 38360148. 

15. Mccammon, Margaret & Pio, Edwina & Barakat, Shima & Vyakarnam, Shai. (2014). Corporate venture capital and Cambridge. Nature biotechnology. 32. 975-978. 10.1038/nbt.3029.

APPENDIX

The process of drug discovery and development

It is broadly divided into two stages—preclinical (nonclinical) and clinical. For the purpose of this manual, the relevant portion is the preclinical phase up to preclinical Proof of Concept (pPoC), which is shown in an expanded form below the main figure.

Explanation of commonly used terms in drug development

What is a druggable target?

A druggable target is a biological molecule, typically a protein, that can bind to a drug with high affinity and whose modulation results in a therapeutic benefit. It must be accessible to drug-like molecules (like small-molecule pills) and actively alter the progression of a disease.

Key Characteristics

To be considered "druggable," a target generally must meet three criteria:

• Disease-Modifying: The target must actively drive the disease, so altering its function will alleviate or cure symptoms.
• Chemically Tractable: The target needs physical "pockets" or binding sites on its surface where a drug molecule can lock in, much like a key in a lock.
• Selective: The drug must be able to affect the target without interfering with other essential bodily functions, minimizing side effects

Common Types of Targets

Most druggable targets fall into a few primary structural categories:

• Receptors: Proteins on the cell surface or inside cells that receive chemical signals (e.g., G-protein-coupled receptors).
• Enzymes: Biological catalysts that speed up chemical reactions in the body (e.g., kinases).
• Ion Channels: Pores in the cell membrane that regulate electrical signals (e.g., calcium channels).
• Transporters: Proteins that carry molecules across cell membranes.

Why the Distinction Matters

While the human body contains thousands of proteins, only a fraction (the "druggable genome") can actually be targeted by current pharmaceuticals. Many disease-causing proteins lack a suitable binding pocket and are considered undruggable. Identifying a druggable target early in target identification minimizes the risk of a drug failing in later development stages.

How does one go about “Selecting an animal model”

This is always a bit of a balancing act. While no animal perfectly mimics a human, the "ideal" model is generally one that mirrors the human condition as closely as possible to ensure that findings actually translate to clinical success.

Here are the key characteristics researchers look for:

1. Biological Similarity (Validity)

• Face Validity: The animal should show the same symptoms or physical characteristics as the human disease (e.g., a mouse with tremors for Parkinson's).
• Construct Validity: The disease should be caused by the same mechanism. If a human disease is genetic, the animal model should have that same genetic mutation.
• Predictive Validity: The model should respond to treatments the same way a human does. If a drug works in the animal, it should have a high likelihood of working in people.

2. Biological Consistency

• Phylogenetic Similarity: Generally, the closer the animal is to humans on the evolutionary tree (like primates or pigs), the more similar their anatomy and physiology will be.
• Genetics: The animal’s genome should be well-mapped and easy to manipulate (which is why CRISPR and mice are such a popular duo).

3. Practicality and Reliability

• Reproducibility: The disease state should be consistent across different labs. If the model is "finicky" or unpredictable, the data won't be reliable.
• Size and Lifecycle: Ideal models are often small, easy to house, and have short lifespans. This allows researchers to study the progression of a disease from birth to old age in a matter of months rather than decades.
• Cost-Effectiveness: High-quality research is expensive, so models that are affordable to maintain and breed are often preferred.

4. Ethical Considerations

• An ideal model should allow for the study of a disease while adhering to the 3 Rs: Replacement (using non-animal methods when possible), Reduction (using fewer animals), and Refinement (minimizing pain or distress).

Of the above considerations that AI spat out, we consider the following to be the most important: Predictive Validity: The model should respond to treatments the same way a human does. If a drug works in the animal, it should have a high likelihood of working in people.

What do we mean by a “Sufficiently large market opportunity”

A large enough marketing opportunity to make developing a new drug worthwhile is generally considered to be a blockbuster potential, defined as a drug that can generate annual sales of US$1 billion or more. Given that the average cost of bringing a new drug to market can exceed $1 billion (ranging from $350 million to over $5 billion), this high-revenue target is necessary to justify the 10-15 year development timeline, failed trials, and high regulatory hurdles.

What do we mean by a “pathway for development/regulatory approval”

All drugs will have to undergo testing in preclinical and clinical phases before approval. If there are examples of previous drugs of the same type approved for the same indication, then the development/regulatory pathway for approval is clear. However, for a first-in-class drug OR a drug for which a surrogate marker of clinical outcome is to be used, both development and regulatory approval can be more complex and require meetings with the regulatory agency in advance where the most suitable development pathway can be discussed and agreed upon in advance. There could be some diseases where it is hard to conduct RCTs (lack of sufficient number of patients or ethical considerations precluding a control group), in which case conditional approval can lead to clinical use along with ongoing monitoring for efficacy and side effects.

What are “guidance documents”

FDA guidance documents explain the agency’s interpretation of, or policy on, a regulatory issue. The FDA prepares guidance documents for regulated industry, its own staff, and the public. FDA uses guidance documents to explain the Agency’s current thinking on such matters as the design, manufacturing, and testing of regulated products; scientific issues; content and evaluation of applications for product approvals; and inspection and enforcement policies.

Although guidance documents are not legally binding, they provide insight to approaches that may help regulated industry reach their regulatory goals. Other approaches that satisfy the relevant law and regulations may be used. FDA periodically reassesses its guidance practices and makes improvements, as appropriate, to ensure that FDA is using current best practices with regard to the prioritization, development, issuance, and use of its guidance documents.

The European Medicines Agency's Committee for Medicinal Products for Human Use prepares scientific guidelines in consultation with regulatory authorities in the European Union (EU) member states to help applicants prepare marketing authorisation applications for human medicines. Guidelines reflect a harmonised approach of the EU Member States and the Agency on how to interpret and apply the requirements for the demonstration of quality, safety, and efficacy set out in the Community directives. 

Target deconvolution: Identifying the target of a drug with an unknown mechanism of action relies on a combination of genetic, biochemical, and computational approaches. Because no single technique is infallible, researchers typically use a tiered strategy.

1. Direct Biochemical Methods

These methods physically isolate the proteins or molecules that your drug binds to within a cell. Affinity Purification (Pull-down Assays): You attach your drug to a solid support (like a bead) and incubate it with cell lysates. Proteins that stick to the drug are washed, separated, and identified using Mass Spectrometry (MS). Photo-Affinity Labeling (PAL): A photoreactive tag is added to the drug. When exposed to UV light, the drug forms a permanent covalent bond with its target, allowing you to identify weak or low-abundance interactions that regular pull-downs might miss.

• Label-Free Target Engagement: Techniques like DARTS (Drug Affinity Responsive Target Stability) or CETSA (Cellular Thermal Shift Assay) rely on the fact that when a drug binds to a protein, it makes the protein more stable and resistant to heat or enzymes.

2. Genetic Approaches

Instead of looking for binding, these methods look at how the drug alters cellular behavior.

CRISPR/Cas9 Screening: Researchers use CRISPR to knock out or knock down specific genes to see which cells become resistant or hyper-sensitive to the drug. If deleting a specific gene makes the cell immune to the drug, that gene's protein is likely the target.

• Overexpression Libraries: You can overexpress various proteins in the cell. If an excess of a specific protein negates the effect of your drug, that protein is probably the intended target.

3. Computational and 'Omics' Approaches

These strategies infer targets based on large datasets and data patterns.

• Transcriptomics (Connectivity Mapping): You treat cells with your drug and sequence the resulting mRNA (gene expression profile). By comparing this profile to databases of known drugs (like the Connectivity Map), you can guess the target if the profile matches a drug with a known mechanism.
• Molecular Docking Simulations: If you already know the 3D structure of proteins involved in your disease pathway, computer algorithms can simulate how your drug docks into their binding sites.

4. Phenotypic Profiling

Sometimes the target isn't a single protein but a pathway. Multi-omics (analyzing metabolic changes or protein expression changes across the whole cell) can map out which signaling networks are altered by the drug, pointing you toward the general area of action.

What is drug-likeness?

Drug-likeness is a qualitative estimate of how likely a molecule is to successfully become an oral drug. It measures a compound's physicochemical and structural similarity to existing, successful pharmaceuticals. Compounds with high drug-likeness generally possess optimal safety and ADMET profiles (absorption, distribution, metabolism, excretion, and toxicity).

Developing a drug is highly challenging because a molecule must survive diverse physiological environments before reaching its target. Evaluating drug-likeness early in the development pipeline prevents researchers from wasting time on compounds that will ultimately fail in clinical trials.

Why Drug-Likeness Matters

• Oral Bioavailability: Ensures the drug can be taken as a pill, survive the acidic environment of the stomach, pass through intestinal walls, and enter the bloodstream.
• Metabolic Stability: Helps the compound resist being destroyed by the liver (first-pass metabolism) before it can exert its therapeutic effect.
• Membrane Permeability: Determines whether a molecule can cross fatty cell membranes to reach its site of action.
• Safety Profile: Reduces the likelihood of toxic accumulations or adverse off-target side effects in the body.

How Drug-Likeness is Evaluated

Scientists use a mix of property-based rules and computational scoring models to assess molecules:

1. Lipinski's Rule of Five (Ro5)

This is the most famous historical guideline. It suggests that an orally active drug usually conforms to the following limits:

• Molecular Weight: Under 500 Da
• Lipophilicity (log P): Less than 5
• Hydrogen Bond Donors: 5 or fewer
• Hydrogen Bond Acceptors: 10 or fewer

2. Veber's Rules

These rules focus on molecular flexibility and surface area. They suggest that a compound is more likely to have good oral bioavailability if it has: [1]

• 10 or fewer rotatable bonds
• A Polar Surface Area (PSA) equal to or less than 140 Angstroms

3. Quantitative Estimate of Drug-likeness (QED)

Because rules like Lipinski's are sometimes too strict (many successful drugs break the rules), modern researchers often use the QED score. The QED takes various physicochemical properties and compresses them into a single, unitless score ranging from 0 (unfavorable) to 1 (highly favorable).

The Evolution of the Concept

While drug-likeness is an excellent guide for small-molecule drug discovery, it is not an absolute rule. Many modern medicines, such as biologics, monoclonal antibodies, and targeted cancer drugs, intentionally fall outside these traditional boundaries. Today, developers combine traditional rule-based filtering with machine learning tools

What is a counterscreen?

These ensure that the drug candidate only affects the intended target and doesn't cause harmful side effects or act in a non-specific way.

The Problem: A compound might bind to the wrong receptor, kill healthy cells (cytotoxicity), or alter unrelated enzymes.

• The Screen: The promising compounds are tested against a panel of similar, unrelated targets or healthy cell lines. If a drug shows activity against these off-target elements, it gets weeded out.

What is preclinical Proof of Concept (pPoC)?

Preclinical proof of concept (PoC) is the initial experimental phase in drug discovery that demonstrates a new therapy, drug, or device has the desired biological effect in non-human models. It provides the foundational evidence that a treatment works before it can be tested in humans.

Key Objectives

The primary goal of preclinical PoC is to de-risk a medical project by answering fundamental scientific questions before investing in expensive, time-consuming clinical trials. This phase validates several critical factors:

Efficacy: Does the drug actually treat the targeted disease or biological pathway in animal or cell models?

• Target Engagement: Does the compound interact with the intended disease target in the body?
• Mechanism of Action: How exactly does the therapeutic intervention produce its beneficial effects?
• Dosing: It helps identify the optimal biological dose, route of administration, and dosing schedule.

Therapeutic Index

The therapeutic index (TI) is a quantitative measurement of drug safety comparing the dose that causes toxicity to the dose that produces the desired effect. A high index means a drug is safe with a wide gap between effective and toxic doses; a low index means the drug is risky and requires careful monitoring. If the No Observed Adverse Effect Dose (or level, if blood concentrations are being measured) is at least 10 times the dose at which the therapeutic benefit is seen, then the TI is 10 and the margin of safety is acceptable. This threshold of acceptability can be different for a drug for a serious (and otherwise untreatable) disease like cancer versus a drug for chronic administration for lowering LDL cholesterol. Benchmarking the TI against available/competing treatments can help determine the acceptability of the TI for further development.

What is developability assessment?

A drug's developability assessment is an early-stage de-risking process that evaluates a molecule’s physicochemical, biophysical, and manufacturing properties. It ensures that a biologically potent candidate can be safely manufactured, scaled, formulated, and administered to patients. Key factors generally focus on the following core areas:

1. Manufacturability & Yield

• Expression & Synthesis: Can the molecule be consistently and cost-effectively produced at high yields in cell culture (for biologics) or via chemical synthesis (for small molecules)?
• Purification & Stability: Is it stable through downstream processing, such as viral inactivation and chromatography?

2. Physicochemical Properties

• Solubility & Permeability: The drug must adequately dissolve in biological fluids and permeate cellular membranes to reach its target. For small molecules, this is often guided by Lipinski's Rule of Five.
• Charge & Isoelectric Point (pI): Determines the drug's charge distribution, which affects formulation pH, solubility, and interaction with manufacturing equipment.
• Hydrophobicity: Excessive hydrophobicity can lead to non-specific binding and precipitation.

3. Stability & Degradation

• Thermal & Mechanical Stress: How well does the drug handle temperature variations (freeze-thaw), shear stress (during pumping/filtration), and prolonged storage?
• Aggregation & Fragmentation: Evaluating a biologic's tendency to clump together or break apart, which reduces efficacy and can trigger immune responses.
• Chemical Degradation: Vulnerability to oxidation, deamidation, or hydrolysis.

Target knockout and knockdown

Gene knockout completely and permanently eliminates a target gene's function at the DNA level. Gene knockdown only partially suppresses expression by destroying its corresponding mRNA, resulting in a temporary reduction of the target protein

Drugs that bind covalently to their targets

Drugs that bind covalently to their targets utilize a mild chemical group (an "electrophilic warhead") that bonds with a nucleophilic amino acid (like cysteine, serine, or lysine) in the target protein. This forms a strong, highly potent, and prolonged interaction.

Common categories and classic examples of covalent drugs include:

1. Targeted Covalent Inhibitors (TCIs)

Engineered for highly specific enzymes, these are predominantly used in cancer and viral therapies to prevent resistance.

• Ibrutinib: Binds to Cysteine-481 on Bruton's tyrosine kinase (BTK) to treat leukemias and lymphomas.
• Osimertinib / Afatinib: Target Cysteine-797 on EGFR to block mutant receptors in non-small cell lung cancer.
• Nirmatrelvir (Paxlovid): Covalently inhibits the main protease of the SARS-CoV-2 virus.
• Rilzabrutinib (marketed under the brand name Wayrilz) is a first-in-class, oral, reversible covalent Bruton's tyrosine kinase (BTK) inhibitor.

2. Classic Irreversible Inhibitors

Some of the most widely used historical drugs rely on covalent binding to permanently inactivate target enzymes. [1]

• Aspirin: Irreversibly acetylates a serine residue on cyclooxygenase (COX) enzymes, shutting down the production of pain and inflammation mediators.
• Penicillin: Covalently attacks and inactivates bacterial transpeptidase (penicillin-binding protein), disrupting bacterial cell wall synthesis.
• Omeprazole: Binds covalently to the hydrogen-potassium ATPase (proton pump) in the stomach, irreversibly blocking acid secretion.

3. Proteasome Inhibitors

• Bortezomib / Carfilzomib: Covalently bind to threonine residues in the catalytic core of the proteasome, disrupting the breakdown of proteins and inducing cell death in multiple myeloma.

Examples of preclinical safety signals that may not translate to human biology

Rodent-Specific Kidney Toxicity (alpha-2u globulin)

1. The Signal: Accumulation of the protein (alpha-2u globulin) in the proximal tubules of male rats, leading to cell death, inflammation, and eventually renal tumors.
2. Human Irrelevance: Humans do not synthesize this protein and hence this pathway poses no carcinogenic risk to humans

Rodent ThyROId Tumors (Enzyme Induction)

• The Signal: Increased incidence of follicular cell tumors in the thyROId glands of rats and mice resulting from chronic administration of xenobiotics.
• Human Irrelevance: In rodents, these are often driven by hepatic microsomal enzyme induction (e.g., via UDP-glucuronosyltransferase) that leads to continuous clearance of thyROId hormones, increased thyROId-stimulating hormone (TSH), and sustained cellular proliferation. This axis is much less prominent in humans.

Peroxisome Proliferation

• The Signal: Proliferation of peroxisomes and related enzymes (like acyl-CoA oxidase) in the liver cells of rodents, which is heavily linked to hepatocarcinogenesis in long-term bioassays.
• Human Irrelevance: Humans possess significantly lower numbers and densities of peroxisomes in hepatocytes and a different molecular receptor (PPAR alpha) threshold. Peroxisome proliferating agents generally do not cause liver cancer in humans.

Species-Specific Hematological & Cardiovascular Variations

• Dog Emesis Reflex: Dogs frequently experience severe emesis and gastROIntestinal upset due to their highly sensitive chemoreceptor trigger zones. While this can indicate toxicity, it often acts as a false positive for systemic safety margins since the reflex threshold is far lower in canines than in humans.
• QT Prolongation in Primates: Transient QT prolongation or cardiac repolarization anomalies in monkeys are sometimes observed during safety pharmacology assays. Because non-human primates have much faster baseline heart rates and differing ion channel configurations, these signals often lack human clinical translation.

Species-Specific Immunogenicity for Biologics

• The Signal: Development of anti-drug antibodies (ADAs) and subsequent immune-complex deposition in standard animal models.
• Human Irrelevance: Human immune system infrastructure and epitope targets are fundamentally different from those of standard preclinical species. Furthermore, because biologics are often engineered for human-specific targets, they may not exhibit pharmacologic activity in animal species.

Why Discrepancies Occur

• Toxicokinetics: Different rates of absorption, distribution, metabolism, and excretion (ADME) mean an animal may process or accumulate a drug differently than humans.
• Physiological Differences: Rodent skin absorbs topical substances significantly faster than human skin, which can artificially inflate local toxicity signals.
• Methodological Misclassification: Because of these false-positive safety signals, many drugs that are successfully and safely used by humans today (e.g., penicillin) historically exhibited fatal toxicities in certain test animals (e.g., guinea pigs).

Structural functionalities that are not desirable in a drug

Functionalities associated with carcinogenicity/mutagenicity: arylamines, ring epoxides, alkane sulfonates, arylnitro functions, azo groups, ring N-oxides and NMe₂ groups, methylols, aliphatic aldehydes, vinyl groups attached to aromatic rings, aziridines, nitrogen mustards and chloramines, benzyl halides, alkylnitrosamines, and alkylurethanes. Such functionalities should be considered less preferred and not as absolutely excludable since there are exceptions, and certain drugs that contain these functionalities have been found to be safe.

A standard 12–15 slide Academic-to-Pharma/Investor Pitch Deck should follow this structure:

1. The Executive Summary

• The Hook: One sentence stating what your discovery is, the modality (e.g., small molecule, biologic), and the target disease indication.
• Value Proposition: Why should the biotech company care? (e.g., First-in-class vs. best-in-class, significant reduction in off-target toxicity).

2. The Unmet Need & Target Rationale

• The Problem: The current standard-of-care (SOC), its limitations (e.g., resistance, side effects), and the resulting disease burden.
• Target Biology: The role of your specific protein, pathway, or receptor in the disease.
• Target Validation: Why is this target actionable? (e.g., genetic knockout studies, biomarker data).

3. The Solution & Mechanism of Action (MoA)

• The Discovery: Introduce the specific drug candidate (e.g., Compound X).
• Mechanism: A clear, high-level summary of how the drug interacts with the target.
• Selectivity/Potency: In vitro assay data demonstrating high specificity and binding affinity.

4. Pre-clinical Data (The Core)

• In Vitro & In Vivo Efficacy: Compelling data demonstrating effectiveness in disease models (e.g., tumor reduction graphs, survival curves, target engagement in tissue)
• Pharmacokinetics (PK): Preliminary data on how the drug is absorbed, distributed, metabolized, and excreted (ADME) in animal models.
• Safety Profile: Any early indication of a favorable therapeutic index (lack of toxicity).

5. Intellectual Property (IP) & Exclusivity

• Patent Status: Composition of matter, method of use, and filing status (provisional, PCT, or granted).
• Inventorship & Ownership: Clearly establish that the IP resides with the university or research institution and is available for licensing.

6. Development Plan & The "Ask"

• Timeline: A realistic Gantt chart showing the path to IND-enabling studies.
• The Partnership Goal: What are you looking for from the biotech company? (e.g., sponsored research agreement, out-licensing for Phase 1, joint venture).

7. Team & Collaborators

• Key Investigators: Highlight the principal investigators, their labs, and relevant experts.
• Advisors: Include clinicians or industry veterans who validate the clinical translation pathway.

Tips for Academic Pitching

• Speak the Industry Language: Biotechs think in terms of value inflection points and developability. Emphasize manufacturability, formulation, and toxicity, not just basic biology.
• Keep the Science High-Level in the Main Deck: Hide complex mechanistic signaling pathways or exhaustive screening data in your "Backup Slides" to use during Q&A.
• Showcase Collaborative Readiness: Prove that you have engaged with your university's Technology Transfer Office (TTO) to facilitate easy contracting.

INCORPORATING A COMPANY/BUILDING A START-UP

The right time to incorporate a company is when a satisfactory Freedom to Operate Search has been completed and has yielded no conflicts and/or the patent has been published. At this point the patent can be transferred to the company under an agreement between the host institution where the academic works and the newly formed company. Sequestering the IP in this manner and giving full rights to the company to exploit the IP enables the company to flexibly conclude licensing deals.

1. Key Triggers for Incorporation

• Intellectual Property (IP) Protection: You should incorporate only after you and your university's Technology Transfer Office (TTO) agree that your invention merits patent protection.
• Exclusive Licensing: Wait to incorporate until you are ready to secure an exclusive, worldwide license from your university for the patent-based technology. Investors require this to fund your spin-out.
• Proof-of-Concept: Data should transcend being merely publication-worthy. You are ready when you have robust, reproducible data that validates a market need and demonstrates a highly de-risked approach to solving an unmet need.

2. When to Seek Funding

The right time to raise capital is right after incorporation. At this stage, your startup is in the pre-seed or seed phase. You can utilize non-dilutive grant funding such as the Small Business Innovation Research (SBIR) or Small Business Technology Transfer (STTR) programs to survive the early "valley of death" before raising major venture capital.

If you are based in India, you can leverage incubation support and seed funding from the Biotechnology Industry Research Assistance Council (BIRAC) to translate academic discoveries into commercially viable products.

3. Assembling the Right Team

Do not wait to incorporate if you are relying on key graduate students or postdocs. Because faculty founders often face time constraints due to academic responsibilities, postdocs and students make the best early-stage contributors to handle proof-of-concept questions and business model validation. However, you will need to bring in a seasoned, business-minded co-founder or CEO to handle fundraising and corporate strategy.

4. Navigating the "Valley of Death"

The early stages of a biotech spinout require navigating the cash-starved gap between discovery and commercialization. Academics should incorporate and begin this journey early, as the entire translational and clinical pipeline can easily take seven to ten years post-spin-out.

Ultimately, perfection doesn't exist, and the opportunity cost of waiting often outweighs the risks. Speak with your institution's entrepreneurship or tech transfer center to assess the commercial potential of your research and determine the ideal window for your spin-out.

Why and where do drugs fail during development?

The high early failure rate explains why pharma is risk-averse for assets in the preclinical phase of development. The high failure rates due to safety signals and poor PK also point to the need to address these issues early on.

Source: Ulrika Kjellgren and 𝘊𝘪𝘵𝘦𝘭𝘪𝘯𝘦 / 𝘕𝘰𝘳𝘴𝘵𝘦𝘭𝘭𝘢 (2014–2023); 𝘉𝘐𝘖 𝘊𝘭𝘪𝘯𝘪𝘤𝘢𝘭 𝘋𝘦𝘷𝘦𝘭𝘰𝘱𝘮𝘦𝘯𝘵 𝘚𝘶𝘤𝘤𝘦𝘴𝘴 𝘙𝘢𝘵𝘦𝘴 (2011–2020); 𝘞𝘰𝘯𝘨 𝘦𝘵 𝘢𝘭., 𝘉𝘪𝘰𝘴𝘵𝘢𝘵𝘪𝘴𝘵𝘪𝘤𝘴 (2018); 𝘏𝘢𝘳𝘳𝘪𝘴𝘰𝘯, 𝘕𝘢𝘵𝘶𝘳𝘦 𝘙𝘦𝘷𝘪𝘦𝘸𝘴 𝘋𝘳𝘶𝘨 𝘋𝘪𝘴𝘤𝘰𝘷𝘦𝘳𝘺 (2016). 

Preclinical proof of concept (pPoC) is a key valuation inflection point in the drug development journey (Ref1, Ref2) and is the point at which corporates show interest in the invention

What are Technology Readiness Levels?

TRLs stand for Technology Readiness Levels. Originally developed by NASA, it is a standardized 9-point scale used to measure the maturity of a technology throughout its research, development, and deployment phases. The table below shows the TRL definitions used by BIRAC.