ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
Time Speaker Project TitleAbstract Zoom Link
2
Monday
8/10/2020
12:00 to 1:00 p.m.		12:00 to 12:15 p.m. Karina HernandezClinical Diagnosis of a Neurological Patient Case All living organisms are composed of cells. Deoxyribonucleic acid (DNA), found within the cell, carries our genetic information. Genetic inheritance is a basic principle of genetics that explains how phenotypes are passed from one generation to the next. The understanding of how genes work is essential to discovering why diseases or genetic variants occur. The goal of clinical diagnostics is to help people who are affected by or at risk for a genetic disorder to live and reproduce as normally as possible. The rationale for this article is the study of a familial case of a patient with Parkinson's disease. This neurological disorder causes progressive degeneration or death of nerve cells. Case Presentation: We report a 32-year-old woman with tremors of the right hand, difficulty standing, problems with coordination, poor balance. The pedigree of the family suggests an autosomal recessive inheritance. Interpretation of DNA variants was performed on annotated variant files after exome sequencing analysis. Variant annotation was conducted with snpEff (v 4.2) and Alamut-Batch (v 1.4.4) based on the RefSeq database. The UCSC Genome Browser (UCSC) was used as an assessment tool. Variants were classified according to the American College of Medical Genetics and Genomics (ACMG) guidelines. Details on the strategy followed for variant interpretation, evaluation of disease-causing candidate variants, disease with pathological mechanism and possible patient management will be presented.https://zoom.us/j/91336135209?pwd=YmdLSFpIem1ZZDJrV2FRV3ZYbUxrdz09
3
12:15 to 12:30 p.m. Linh NguyenCharacterization of A Genetic Suppressor of A Histone H3 Mutation that Perturbs Transcription Elongation in Budding Yeast.The basic unit of chromatin, the nucleosome is a protein/DNA complex consisting of 146 base pairs of DNA wrapped around a histone octamer. Nucleosomes are targets of regulation for many biological processes that function on DNA, such as gene transcription. We are interested in how histone H3, one of the four core histone proteins, regulates the structure and function of chromatin during transcription elongation. Our lab has previously identified histone H3 mutations which we hypothesize to disrupt chromatin structure, making chromatin more permissive to transcription elongation. We isolated genetic suppressor mutations of one of these H3 mutations, hht2-15R. We hypothesize that this suppressor mutation restores normal chromatin structure in our H3 mutant. Whole-genome sequencing of a yeast strain carrying one of these suppressors revealed mutations in two genes; IPP1, which encodes an inorganic phosphatase, and PEP12, which encodes a t-snare protein involved in vacuole functions. However, genetic analysis shows that the suppressor falls in a single gene. We performed genetic linkage analysis to determine if either PEP12 or IPP1 is linked to the H3 suppressor mutation. Surprisingly, our data revealed that pep12-knockout and hht2-15R have synthetic lethal interaction. We hypothesize that vacuole functions are essential for cells’ viability when chromatin structure are altered, and point-mutations in PEP12 might give rise to PEP12 nuclear function. The vacuole’s functions have not been previously implicated in being essential for cell’s viability. Thus, we are working to test our hypothesis and understand how the vacuole functions are essential when chromatin structure is altered.https://ucsc.zoom.us/j/7131347381
4
12:30 to 12:45 p.m. Shamari WallerAnalysis of NOTCH2NL Gene Variants and CNVs in Autism CohortsCharacterizing the variations in human lineage-specific genes responsible for brain development and expansion has posed a challenge in the field of genomics. Advancements in sequencing technology has allowed us to identify specific regions in the genome that can provide insight into neurodevelopmental disorders. Our research investigates NOTCH2NL gene and its role in brain expansion and human cortical development in human evolution. NOTCH2NL, found in the Q21 region on chromosome 1, is in a location in the human genome that undergoes distal duplication syndrome. This syndrome has increased the amount of varied copies of NOTCH2NL humans have, affecting the signaling of NOTCH receptors responsible for brain/intellect development. Within the NOTCH2NL region we explore that there may be an inherited variant activating the Autism Spectrum Disorder phenotype. Using clinical data from Simons Research Initiative, we align patient sequence data to the hg38 human reference genome and a NOTCH2NL consensus genome. The alignments will reproduce the sequence data in a variant call format (VCF) outputting a list of any gene variants that may arise. The VCF will help produce singly unique nucleotide (SUN) identifier maps that graph the frequency of a variant at each base position within the NOTCH2NL gene. We can then filter the cohort for variants in >1% of the population (dbsnp) and not known to be benign (clinvar), leaving us with locations of unique variants. After we have characterized unique variants and can identify their locations we can better support our exploration of Autism Spectrum Disorder phenotype in a variant in the NOTCH2NL region.https://ucsc.zoom.us/j/2961986133
5
12:45 to 1:00 p.m. Teresa MarquezClinical diagnosis of hereditary diffuse gastric cancer through phenotypic and genomic data analysis.Current cancer diagnostic tools, a combination of genomic and phenotypic data analysis, are used to characterize gene variants reported to cause susceptibility. For example, our project investigates a patient whose family history of stomach cancer led us to believe that her cancer diagnosis was also stomach-related. However, the presence of stomach cancer usually involves a mass being found, which was not the case with this patient. This prompted a genetic analysis resulting in two candidate variants being detected within the patient's CDH1 and MLH1 genes. It has been reported that mutations in the germ line of the CDH1 gene are known to cause hereditary diffuse gastric cancer (HDGC) (Pharoah et al., 2001). This cancer is known to cause a thickening of the stomach wall without producing an actual mass (Kaurah & Huntsman, 2002). This project will characterize diagnostic and treatment solutions through phenotypic and genomic clinical data analysis of this HDGC patient. Using a combination of genomic databases, we will be reviewing the patient’s phenotype and cross checking a list of possible pathogenic gene variants. Preliminary results will define the severity of diagnostic markers, the pathological mechanism of the disease, the disease-gene validity, and potential medical solutions for treating the disease.https://csumb.zoom.us/j/95606683817
6
7
Tuesday
8/11/2020
12:00 - 1:30 p.m.		12:00 to 12:15 p.m. Valerie YaceFinding Evolution-Imposed Geometries in Cancer Genomics Data to Identify New Patient Subtypes of Diseasehttps://us02web.zoom.us/j/86814487676?pwd=SVk1ZjJ2L0FZR2t0VnUxakR6QWVBZz09

Meeting ID: 868 1448 7676
Passcode: 7ic9RN
8
12:15 to 12:30 p.m. Samantha MillerEvolutionary Analysis of the CTCF GeneCTCF, or CCCTC-binding factor, is a protein encoding gene that is highly conserved across metazoa. The CTCF protein is an eleven-zinc-finger protein that regulates transcription as a repressor and activator, functions as an enhancer-blocking insulator, and contributes to genomic organization by establishing and maintaining chromatin loops. While CTCF is a core regulatory gene that exercises crucial control over genome-wide transcriptional activity, its evolutionary history remains poorly understood. To elucidate the functional importance of evolutionarily conserved regions of the CTCF gene, we collected transcripts of CTCF orthologs from 65 bilaterian species with publicly accessible Hi-C sequencing data. We aligned the transcripts and generated a maximum likelihood phylogenetic tree based upon the alignment. The tree displayed four distinct clades of mammals, insects, fish, and aviators and reptiles. Within the alignment, we identified a range that consists of 932 nucleotide bases conserved across all species of analysis with frequent synonymous mutations in wobble base positions. We observed a 44-base deletion of otherwise conserved sequence specific to the insect clade. By constructing predicted protein structures, we determined that the deletion encodes the third C2H2 zinc finger protein domain in species in which the sequence is still present. These findings will be used to inform a study investigating nuclear spatial structural data for the species included in our analysis that will identify correlation between systematic changes in the spatial dataset with changes in our sequence alignment. https://csumb.zoom.us/j/95384049807
9
12:30 p.m. to 12:45 p.m.Gifti GemedaUsing a Notch-reporter assay to understand the ability of specific NOTCH2NL alleles to activate Notch signalingThe main goal is to understand how the human specific gene NOTCH2NL affects neural development. Duplications and deletions in the region 1q21.1, which contains three NOTCH2NL genes, are associated with neurodevelopmental phenotypes such as microcephaly, macrocephaly, autism and schizophrenia. Notch signaling is essential for brain development, determining the timing and duration of neural progenitor proliferation and neural differentiation. The goal of this project is to understand the function of the various NOTCH2NL genes and alleles, specifically their ability to activate the Notch pathway. Previous work has identified a surprising variety of alleles in healthy populations. Site directed mutagenesis and plasmid cloning is used to make the different alleles of NOTCH2NL. Previous work testing a small subset of these alleles has shown them to have varying potencies to enhance Notch signaling and interact with NOTCH receptors. We are using a CSL-Luciferase based Notch reporter assay to measure the ability of individual NOTCH2NL alleles to affect notch pathway activity. To determine the spectrum of NOTCH2NL alleles in the human population, Molecular Inversion Probes (MIPs) along with Illumina sequencing are used to assess the NOTCH2NL alleles present in a large number of individual genomes. By understanding the variation of NOTCH2NL in the general and affected population we will better understand the association of genotype and phenotype of NOTCH2NL and neurodevelopmental disorders.https://ucsc.zoom.us/j/96407331733
10
11
Wednesday
8/12/2020
12:00 - 1:30 p.m. 		12:00 to 12:15 p.m. Gavin DrummClassifying variants in patients with neurodegenerative disordersNeurodegenerative disorders occur through the progressive decay and loss of motor neurons due to protein malformation. Developments in sequencing technology is allowing us to characterize the genes responsible for downregulation in cellular processes leading to progression of diseases. In order to classify pathogenic variants in the genome, we will be using six patient cases and ranking their susceptibility to neurodegenerative disorders based on their respective variants and phenotypes. We will be using a combination of genome browsers, a phenotype ontology database, and a mendelian trait database to define susceptibility. Using the Genome Aggregation Database (gnomAD) as a reference control, we will cross-reference and analyze each patient’s gene-phenotype relationships and variants through the Human Phenotype Ontology (HPO) database.This analysis will be further filtered with the American College of Medical Genetics (ACMG) guidelines for evaluating the identified variants as benign or pathogenic, allowing us to link each pathogenic variant to a particular disorder. Using this holistic approach, we will be able to uncover the relationship between variants and how their presence may lead to an array of aberrant neurodegenerative disorders.https://csumb.zoom.us/j/98256436039
12
12:15 to 12:30 p.m. Timothy HannemanDiplotyping human genomes with long nanopore readsGenomic variation detection, also known as variant calling is the process of finding small changes in individuals as compared to a reference genome. Oxford nanopore sequencing technology (ONT) can produce long but erroneous reads of genomic sequences. Though erroneous, long reads are able to align to the reference genome with high confidence and can give resolution to the most challenging regions of the genome. However, because the reads have approximately ~7-10% error rate, variant identification is difficult. Recently, a variant calling pipeline based on Deep Neural Networks (PEPPER-DeepVariant) showed that ONT based variant calling could achieve high-quality single-nucleotide polymorphism (SNP) results. The pipeline uses a recurrent neural network (RNN) based polisher (PEPPER) to find potential variants and a convolutional neural network-based genotyper (DeepVariant) to infer which potential candidates are errors and which are true variants. Utilizing a GPU high-accuracy run on HG002 chr20 sample data, the PEPPER variant calling pipeline can accurately detect SNP variants with a precision and recall of greater than 0.99 with ~50x coverage. To improve the pipeline, we benchmark different recurrent neural network models. This work will focus on improving the pipeline's generalizability by evaluating different RNNs for candidate variant identification.https://csumb.zoom.us/j/99285653935
13
12:30 to 12:45 p.m. Ibrahim JabarkhelBuilding Docker-based Reproducible Workflows in BioinformaticsThe genomic scientific community is producing enormous amounts of genomic data. Cloud-based systems and other computing resources are required in order to store and analyze the genomic data faster and proficiently. The current reproducibility crisis in science requires reproducibility technology to not be overburdened by the data. Docker is a reproducibility tool that is environment-agnostic that containerizes an application with all the required libraries and dependencies. Our research will reward the scientific community of Bioinformatics with more transparent and reproducible tools built with Docker and the Workflow Description Language (WDL). We are using Docker to create images and containers and saving them on Dockstore.org for the public to access these pipelines. Our docker with a descriptor workflow is helping researchers by turning any genomic data into subsets to examine and test their work easily. Researchers can use our reproducible workflows to learn, create, and experiment with their own genomic analysis workflows.https://zoom.us/j/9860163373?pwd=YUhvR2x3Yk4xcmxzd1hFemd0N1BMdz09
14
12:45 to 1:00 p.m. Nicholas VasquezBuilding Containerized Workflows in BioinformaticsIn an effort to make the process of analyzing genomic data faster and more efficient, many scientific processes are now being executed through the cloud. As new methods of working with large datasets through cloud based systems are developed, researchers must make sure that these methods do not contribute to the reproducibility crisis in science. Docker is a technology that packages all libraries and dependencies required for a data analysis project and deploys it on any computing environment. Our work involves taking analysis pipelines of genomic data, translating them into docker-based reproducible workflows and publishing them on Dockstore.org. Dockstore has features for reproducibility including links to original source code on Github, links to Docker containers, and descriptor languages that make source code easily executable in any environment. Our project involves using Docker and the Workflow Description Language (WDL) to write workflows that subset data from large genomic datasets which can be used for teaching and testing purposes. The reproducible workflows we create will provide researchers with new resources to learn, build and test their own genomic analysis workflows.https://csumb.zoom.us/j/4592565831
15
16
All meetings are delayed today by 10 minutes. Thank you for your patience!
17
Thursday
8/13/2020
12:00 - 1:30 p.m. 		12:05 to 12:20 p.m. Héctor CarriónTowards Classification of Wound Stages Using Deep Learning AlgorithmsWound healing is a variable, complicated, painful, and often prolonged process that involves a series of different stages. Current healing solutions are passive in nature and do not account for the dynamic state of the wound. A proactive treatment solution could benefit extensively from real-time awareness of the different healing stages (i.e., what physiological process has just completed, what process is in progress, and what process should subsequently take place). Therefore, to enable such proactive treatments, our research focuses on an end-to-end deep learning pipeline that intakes a wound image and returns its size, healing stage, and other critical parameters related to the injury. We also aim to study the changes of wound conditions over time and integrate those findings back onto the healing solution. We hypothesize that well-known object detection and medical instance segmentation deep learning algorithms could be adapted to our particular task. Our preliminary results after developing customized algorithms based on the YOLO, U-Net, and DenseNet architectures are promising and can already locate and segment the wounded area. We also plan on exploring unsupervised methods and comparing their performances. Moreover, after post-processing our results, we expect to build a clear end-to-end graphical representation of the healing process. Ultimately, the results of this project should be an accessible, easy to implement, and robust wound monitoring Artificial Intelligence (AI)-based pipeline that will provide useful, actionable insights geared towards proactive wound treatments.https://zoom.us/j/8748393793?pwd=WHg2RDFUeU9XalNEbi94eCtndmpQQT09

Meeting ID: 874 839 3793
Passcode: 0FMVtz
18
12:20 to 12:35 p.m. Christina MaganaData visualization of cancer variant impactsSomatic mutations in cells are known to be major contributors to cancer. They can manipulate or alter the function of genes having a domino effect because of the sensitive ecosystem of the human body. Advancements in high-throughput sequencing of tumor cells have identified millions of somatic mutations. While our ability to identify these mutations has increased their implications are mostly unknown. The identification of a mutation’s functional impact is a key step for both individual patients and the general population. Individual patients benefit through more specialized treatments. While the general population benefits through mutations of high impact being red-flagged. A python tool was developed to predict the functional impact of somatic mutations, called expression-based variant impact phenotyping (eVIP). eVIP uses overexpression of wild-type and mutant forms of a gene in cell lines and profile gene expression to perform its analysis. The mutations are assessed as gain-of-function, loss-of-function, change-of-function, neutral, and non-informative. A modification to eVIP led to eVIP2 which uses RNA sequencing versus L1000. The difference between the two is that L1000 is an assay that measures mRNA transcript for around 1000 genes (978 genes) while RNA sequencing is an assay of the entire transcriptome. eVIP2 uses RNA sequencing to analyze the whole transcriptome in order to assess the pathway level consequences. eVIP2 outputs multiple graphs and files per mutation, therefore when analyzing multiple mutations simultaneously the output can be overwhelming. Our goal is to create a summary report for eVIP2 to summarize the results and make it easier to interpret. In particular, our plan is to make the summary report available in an HTML report or website report where tables, graphs, and links to other folders will be found. There will also be a help page that will assist with the interpretation of the figures. To evaluate the effectiveness of the HTML report we will use a data set of two variants of the protein-coding gene RNF43 as a test case. Moreover, mutations of the gene have frequently been detected in colorectal, gastric, endometrial cancer. Utilization of eVIP2 for clinical research may give rise to specialized treatment for patients with somatic mutations. The summary report would make it more feasible for researchers and/or clinicians to implement the software in their practices. https://csumb.zoom.us/j/91778558065
19
12:40 to 12:55 p.m. Asgton StephensonClinical Diagnosis of Patient's Exome Sequencing Unveils Variant in ABCA4 Gene Causing Retinitis Pigmentosa DisorderRetinitis Pigmentosa (RP) is a group of related eye disorders affecting the retina, and characterized by progressive vision loss (Bethesda, 2020). The vision loss is attributed to the degeneration of the eye’s specialized light receptor cells, known as photoreceptors. The case of the patient comes from a consanguineous family with a prevalent history of visual field impairments is reviewed. The patient’s phenotype is cross referenced with numerous disease and genomic databases, leading to identification of the pathologic variants found. Two candidate variants were identified, both presenting as pathogenic and non-synonymous within the ABCA4 gene, in accordance to American College of Medical Genetics (ACMG) guidelines. Animal models and previous findings suggest that the mutations within the related ABCR gene, can be associated phenotypically with retinal diseases (Shroyer et al., 1999). This project design of this case will distinguish the patient’s respective clinical diagnosis and suggest treatment options or disease management strategies to complete the clinical interpretation.
https://csumb.zoom.us/j/91877868352
20
12:55 to 1:10 p.m. Rosa SanchezAnalysis of the gene and protein sequences shows the relationship between organisms that contain the Ars OperonArsenicals are toxic compounds that interfere with ATP synthesis, protein function, and cell growth. Removing these toxic compounds from the cell can be done through methylation, an As(III) specific transporter, and the reduction of As(V) to As(III). For the latter, the reduction of As(V) to As(III) is mediated by ArsC, the arsenate reductase coded for in the ars operon. Our model organism, Citrobacter sp. TSA-1, was isolated from termite hindguts and codes for 3 arsC’s, one in the ars operon and two other arsCs located in other parts of the genome. In previous research, deletion of the arsC in the ars operon showed that arsenate reduction was still possible, though at lower arsenate concentrations. We hypothesise that two of the three ArsC’s functions as an arsenate reductase and are evolutionarily conserved, while the other ArsC does not as it is truncated and lacks key arsenite binding residues. To test our hypothesis, we used NCBl to retrieve our DNA and protein sequences, Clustal Omega to align these sequences, Phylogeny.fr to establish a phylogenetic tree, and figtree to edit our tree. If preliminary results demonstrate the conservation of arsenate reductase function, then we may predict that Citrobacter sp. TSA-1 encodes various arsenate reductases crucial to cellular removal of toxic compounds.https://cccconfer.zoom.us/j/7788403514
21
1:10 to 1:25 p.m. Serafina NievesIdentifying and validating horizontal gene transfer events from Wolbachia to arthropodsWolbachia are gram-negative bacteria that frequently infect arthropods and form endosymbiotic relationships with the host (Werren 1997). Due to the close proximity of Wolbachia endosymbiont genomes and arthropod host genomes, horizontal gene transfer (HGT) has been shown to occur from Wolbachia to the host in several organisms, including mosquitoes, flies, aphids, and wasps (Hotopp 2011). However, studies documenting the presence of Wolbachia HGT in arthropod genomes are limited to a few species that have shown extensive transfer. We developed a high-throughput bioinformatics workflow that can identify and validate putative HGT events using publicly available sequence data. Using our comprehensive screening process, we found 1673 putative horizontally transferred protein coding sequences across 16 species, most of which do not have documented Wolbachia-to-arthropod HGT events. We also developed a method to validate these Wolbachia-to-host insertions using available whole genome sequencing data to compare the depth of coverage of putative Wolbachia insertions to coverage of exons in the host genome. We found that Wolbachia-to-arthropod HGT events may be more common than current literature, or lack thereof, seems to suggest.https://ucsc.zoom.us/j/93107330141
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100