1 of 37

Emerging Genetic Diagnostic Approaches at the Mount Sinai Center for Undiagnosed Diseases

New York State Genetics Task Force

6/3/26

Louise Bier, MS, CGC

Rachel Evard, MS, CGC

2 of 37

Conflicts of Interest

The presenters have no conflicts of interest to disclose.

3 of 37

Learning Objectives

Understand new technological approaches that can help to arrive at clinical genetic diagnoses
Explain the clinical, diagnostic, and support resources available at the Center

3

4 of 37

4

Undiagnosed Diseases Network Diagnostic Center of Excellence �Icahn School of Medicine at Mount Sinai

Principal Investigators

Bruce Gelb, MD

Manisha Balwani, MD, MS

Co-Investigators

Mafalda Barbosa, MD, PhD

Louise Bier,

MS, CGC

Dusan Bogunovic, PhD

Charlotte Cunningham-Rundles, MD, PhD

Carol Horowitz, MD, MPH

Ayuko Iverson, MD

Site Coordinator

Rachel Evard, MS, CGC

Bioinformatician

Mariya Shadrina,

PhD, MS

Community Partners

Eric Gayle, MD

CEO of IFH, Co-PI

Saskia Shuman, MHS, PhD, Co-PI

Joanna Jen, MD, PhD

Rory Abrams, MD

Julie Schoonover, MD

Clinical Champion, Co-I

5 of 37

Program Overview

Undiagnosed Diseases Network is a research study funded by the National Institutes of Health
Brings together clinical and research experts from across the United States to solve the most challenging medical mysteries using advanced technologies

5

Program Goals and Objectives

Improve level of diagnosis and care for patients with undiagnosed diseases
Facilitate research into etiology of undiagnosed diseases
Create an integrated and collaborative research community to identify improved options for optimal patient management

6 of 37

Clinical Site Research Core Data Management Coordinating Center Sequencing Core

7 of 37

Long-Read Genome Sequencing

7

8 of 37

Long-Read Genome Sequencing

Uses DNA fragments from 1,000 – 20,000+ bases (vs. traditional sequencing 50-300 base fragments)

Allows for better detection of structural variation (e.g., deletions, insertions, inversions, duplications, complex rearrangements)
Allows for detection of repeat expansions
Improves resolution of complex genomic regions, including pseudogenes and highly homologous sequences
Phasing

8

Logsdon GA, Vollger MR, Eichler EE. Long-read human genome sequencing and its applications. Nat Rev Genet. 2020;21(10):597-614. doi:10.1038/s41576-020-0236-x

9 of 37

Alignment to Reference Genome

9

Reference Genome

AGCTGATTTCGACTGACTTACAGGGAGGAGGAGGAGGAGGAGGATACGTAGC

GGAGGAGGA

Key Takeaway: Read length impacts accuracy of alignment to reference genome

TTACAGGGAGGAGGAGGAGGAGGAGGATAC

Short Reads: ~50-300 base pairs

Long Reads: ~1,000-20,000+ base pairs

10 of 37

Diagnostic Yield

Adds diagnoses in ~7-17% of cases after negative short-read genome sequencing

Most additional diagnoses driven by:

Structural variants not resolved by short-read sequencing
Repeat expansion disorders
Improved resolution of complex/homologous genomic regions

10

Del Gobbo GF, Boycott KM. The additional diagnostic yield of long-read sequencing in undiagnosed rare diseases. Genome Res. 2025 Apr 14;35(4):559-571. doi: 10.1101/gr.279970.124. PMID: 39900460; PMCID: PMC12047273.

11 of 37

Long-Read Genome Sequencing Case Example: Overview

61-year-old woman with refractory adult-onset epilepsy and cortical tremor

First generalized tonic-clonic seizure at 38, stimulus-sensitive
Family history of adult-onset myoclonic seizures and progressive movement disorders (father, brother, paternal uncle & his daughter)

Consistent with Familial Adult Myoclonic Epilepsy – associated with multiple different genes

11

Perera BLA, Stewart R, Furuta Y, et al. First clinical diagnosis of FAME3 via commercial Long-Read sequencing reveals mosaic repeat expansion in MARCHF6 gene. Neurogenetics. 2025;26(1):61. Published 2025 Aug 11. doi:10.1007/s10048-025-00835-6

Prior Testing (beginning 2016)

Myoclonic epilepsy panel = negative
Expanded epilepsy panel = negative

Sequencing and del/dup

Genome sequencing (i.e. short-read) = negative

Long-read genome sequencing (2024)

12 of 37

Long-Read Genome Sequencing Case Example: Results

Pathogenic (TTTTA/TTTCA)n repeat expansion observed in MARCHF6 gene

12

Perera BLA, Stewart R, Furuta Y, et al. First clinical diagnosis of FAME3 via commercial Long-Read sequencing reveals mosaic repeat expansion in MARCHF6 gene. Neurogenetics. 2025;26(1):61. Published 2025 Aug 11. doi:10.1007/s10048-025-00835-6

13 of 37

Long-Read Genome Sequencing Case Example: Results

Pathogenic (TTTTA/TTTCA)n repeat expansion observed in MARCHF6 gene

13

Perera BLA, Stewart R, Furuta Y, et al. First clinical diagnosis of FAME3 via commercial Long-Read sequencing reveals mosaic repeat expansion in MARCHF6 gene. Neurogenetics. 2025;26(1):61. Published 2025 Aug 11. doi:10.1007/s10048-025-00835-6

Key Limitation:

Intronic, repeat expansion

14 of 37

RNA Sequencing

14

15 of 37

RNA Sequencing

A method for sequencing an entire set of RNA molecules (transcriptome) in a tissue sample

Detects:

Changes in RNA sequence
Changes in RNA amount (expression)

RNA-seq + exome/genome increases yield by 10-35% over exome/genome alone

15

Zhao S, Macakova K, Sinson JC, et al. Clinical validation of RNA sequencing for Mendelian disorder diagnostics. Am J Hum Genet. 2025;112(4):779-792. doi:10.1016/j.ajhg.2025.02.006

RNA-seq is a method of sequencing an entire set of RNA molecules, or the transcriptome, in a tissue sample. RNA Sequencing can give us two important pieces of information: it can detect changes in the RNA sequence itself, giving us information on what the final protein chain will be – which means we can better understand if a variant impacts how the introns and exons are spliced. Over on the right, you can see some examples of aberrant splicing, such as exon skipping or intron retention.

RNA sequencing can also tell us how much a gene is expressed – so rather than looking at the RNA sequence itself, we can see whether there is more or less of the gene expressed compared to normal. In the graph on the left, you can see an example of aberrant gene expression identified through RNA sequencing analysis.

So RNA sequencing can be used alongside exome or genome sequencing in order to better interpret variants of uncertain significance by revealing their effects on splicing and gene expression. Studies have shown that this integration can increase diagnostic yield by 10-35%, depending on tissue source and indication.

16 of 37

RNA Sequencing Case Example: Overview

Sibling probands evaluated at Baylor College of Medicine's UDN clinical site with a clinical diagnosis of Fine-Lubinsky syndrome:

Brachycephaly
Hearing loss
Cataracts
Microstomia
CNS anomalies
Developmental delay

Syndrome first described in 1983 but no known genetic cause until 2024

16

Waskow ER; Undiagnosed Diseases Network, Emrick LT, et al. Recessive loss-of-function variants in DPH1 identified as the molecular cause in a sibling pair previously diagnosed with Fine-Lubinsky syndrome. Am J Med Genet A. 2025;197(1):e63845. doi:10.1002/ajmg.a.63845

Prior Testing

Chromosomal microarray analysis
Quad exome sequencing (2 affected sibs + both parents) = negative

UDN Evaluation

Research analysis of exome data identified homozygous intronic variants in DPH1 in BOTH siblings (c.749+39G>A)

Parents found to be heterozygous for the variant
Unaffected sibling not found to be a carrier

RNA sequencing

17 of 37

RNASeq Case Example: RNASeq Analysis

17

RNASeq analysis of Proband 2 fibroblasts revealed:

Significantly reduced expression of DPH1 compared to the controls

Waskow ER; Undiagnosed Diseases Network, Emrick LT, et al. Recessive loss-of-function variants in DPH1 identified as the molecular cause in a sibling pair previously diagnosed with Fine-Lubinsky syndrome. Am J Med Genet A. 2025;197(1):e63845. doi:10.1002/ajmg.a.63845

18 of 37

RNASeq Case Example: RNASeq Analysis

18

RNASeq analysis of Proband 2 fibroblasts revealed:

Significantly reduced expression of DPH1 compared to the controls
Difference in splicing between proband 2 and the controls

Waskow ER; Undiagnosed Diseases Network, Emrick LT, et al. Recessive loss-of-function variants in DPH1 identified as the molecular cause in a sibling pair previously diagnosed with Fine-Lubinsky syndrome. Am J Med Genet A. 2025;197(1):e63845. doi:10.1002/ajmg.a.63845

19 of 37

RNASeq Case Example: RNASeq Analysis

19

RNASeq analysis of Proband 2 fibroblasts revealed:

Significantly reduced expression of DPH1 compared to the controls
Difference in splicing between proband 2 and the controls

Waskow ER; Undiagnosed Diseases Network, Emrick LT, et al. Recessive loss-of-function variants in DPH1 identified as the molecular cause in a sibling pair previously diagnosed with Fine-Lubinsky syndrome. Am J Med Genet A. 2025;197(1):e63845. doi:10.1002/ajmg.a.63845

Diagnosis: DEDSSH1 (developmental delay with short stature, dysmorphic facial features, and sparse hair 1)

20 of 37

Optical Genome Mapping

20

21 of 37

Optical Genome Mapping

21

Smith, A. C., Neveling, K., & Kanagal-Shamanna, R. (2022). Optical genome mapping for structural variation analysis in hematologic malignancies. American journal of hematology, 97(7), 975–982. https://doi .org/10.1002 /ajh.26587

Can detect:

Structural variants (e.g., deletions, insertions, duplications, inversions, translocations)
Copy number variants
Aneuploidies
Absence of heterozygosity
Repeat expansions and contractions
Complex chromosomal rearrangements

Optical Genome Mapping is an emerging cytogenetic technology that detects structural variants and copy number variants across the genome at much higher resolution than traditional karyotyping.

In the workflow, high-molecular-weight DNA is isolated from patient cells. High-molecular-weight DNA is very long, intact strands of DNA typically around 20 kilobases or longer. This DNA is then fluorescently labeled at a six base pair motif which occurs on average every 5 kilobases, and loaded onto a chip, where nanochannels linearize the DNA molecules for imaging. The images are converted into digital molecules and analyzed using bioinformatic pipelines to detect structural variation. In the circos plot on the far right, the genome is displayed in a circular layout, making it easier to visualize interactions between distant chromosomes. Here, the connecting line across the center demonstrates a translocation between chromosomes 8 and 21.

Optical genome mapping can detect a wide range of genomic variation in a single assay, including deletions, insertions, duplications, inversions, translocations (both balanced and unbalanced), copy number gains and losses, aneuploidies, absence of heterozygosity, tandem repeat expansions and contractions, and complex rearrangements. This comprehensive detection capability allows OGM to capture structural variation that may be missed when relying on any single standard-of-care cytogenetic technique alone.

22 of 37

Comparison to Standard of Care

22

Levy, B., Burnside, R. D., & Akkari, Y. (2025). Optical Genome Mapping: A New Tool for Cytogenomic Analysis. Genes, 16(8), 924. https://doi .org/10.3390 /genes16080924

23 of 37

Optical Genome Mapping Case Example: Overview

4-year-old boy with an epileptic encephalopathy of undiagnosed molecular origin

Refractory epilepsy, severe hypotonia, cortical visual impairment and severe global developmental delay
Seizures began at age one month and consisted of infantile spasms and tonic-clonic seizures

23

Cope, H., Barseghyan, H., Bhattacharya, S., Fu, Y., Hoppman, N., Marcou, C., Walley, N., Rehder, C., Deak, K., Alkelai, A., Undiagnosed Diseases Network, Vilain, E., & Shashi, V. (2021). Detection of a mosaic CDKL5 deletion and inversion by optical genome mapping ends an exhaustive diagnostic odyssey. Molecular genetics & genomic medicine, 9(7), e1665. https://doi .org/10.1002 /mgg3.1665

Prior Negative Testing

Karyotype
Chromosomal microarray
Infantile epilepsy panel (2013)
Mitochondrial genome sequencing (2014)
Trio exome sequencing (2014)
Trio genome sequencing

Optical Genome Mapping

24 of 37

Optical Genome Mapping Case Example: Results

24

Cope, H., Barseghyan, H., Bhattacharya, S., Fu, Y., Hoppman, N., Marcou, C., Walley, N., Rehder, C., Deak, K., Alkelai, A., Undiagnosed Diseases Network, Vilain, E., & Shashi, V. (2021). Detection of a mosaic CDKL5 deletion and inversion by optical genome mapping ends an exhaustive diagnostic odyssey. Molecular genetics & genomic medicine, 9(7), e1665. https://doi .org/10.1002 /mgg3.1665

Using optical genome mapping, researchers were able to visualize large-scale genomic architecture directly. OGM detected a mosaic 90 kb deletion and inversion affecting the CDKL5 gene, which is found on the X chromosome. Just to orient you, given this was a mosaic variant, there are two maps of the proband's x chromosome shown in yellow with the typical Map 1 shown above the blue reference genome and the Map 2 with the complex structural variant shown below the reference genome. The deletion removes a segment of the CKDL5 gene

The box at the bottom demonstrates that no parental DNA molecules spanned proband map 2, suggesting that the complex rearrangement was de novo.

Although CMA is typically able to identify deletions at this resolution, given the mosaic nature of the deletion, it was not identified on CMA. Additionally, even if the rearrangement was not mosaic, CMA would not have been able to fully characterize the complexity of the structural variant due to the balanced nature of the inversion.

This complex rearrangement detected on OGM led to a diagnosis of X-linked dominant early infantile epileptic encephalopathy Type 2 caused by deficiency of CDKL5.

***the paper does not mention the level of mosaicism

25 of 37

Optical Genome Mapping Case Example: Results

25

Cope, H., Barseghyan, H., Bhattacharya, S., Fu, Y., Hoppman, N., Marcou, C., Walley, N., Rehder, C., Deak, K., Alkelai, A., Undiagnosed Diseases Network, Vilain, E., & Shashi, V. (2021). Detection of a mosaic CDKL5 deletion and inversion by optical genome mapping ends an exhaustive diagnostic odyssey. Molecular genetics & genomic medicine, 9(7), e1665. https://doi .org/10.1002 /mgg3.1665

Deficiency of CDKL5

X-linked Dominant Early Infantile Epileptic Encephalopathy Type 2

26 of 37

Animal Models

26

27 of 37

Animal Models

Functional studies investigate variants/genes of uncertain significance

Knockout (gene is made inactive) vs. Knockdown (gene activity is reduced vs. Knockin (altering or inserting gene sequence)

Animal models:

Many different species may be considered – mice, fruit fly (drosophila), zebrafish, worms, etc.

27

28 of 37

Animal Model Case Example: Overview

28

Patient #1: 6yo girl with growth & developmental delay

de novo variant in SRRM1 c.1503del,

p.(Ser502Glnfs*11)

Patient #2: 11yo girl with developmental delay, mild intellectual disability, mild short stature

de novo variant in SRRM1

c.1964C>G, p.(Ser655*)

Patient #3: 17yo boy with autism, significant intellectual disability

de novo variant in SRRM1

c.163A>T, p.(Arg55*)

Submitted to GeneMatcher

Altay MF, Gregor A, Braun D, et al. Heterozygous loss of SRRM1 may be associated with neurodevelopmental phenotypes and anomalies in cell growth and neurite morphology. Eur J Hum Genet. Published online October 27, 2025. doi:10.1038/s41431-025-01966-y

29 of 37

Animal Model Case Example: Animal Model Results

29

Srrm1 knockdown

Expression reduced 30%

Expression reduced 50%

30 of 37

Animal Model Case Example: Animal Model Results

30

Srrm1 knockdown

Expression reduced 30%

Expression reduced 50%

Increased spasms among knockdown flies after stimulus

31 of 37

Animal Model Case Example: Animal Model Results

31

Srrm1 knockdown

Expression reduced 30%

Expression reduced 50%

Decreased climbing among knockdown flies after stimulus

32 of 37

Animal Model Case Example: Overall Results

32

Strong evidence for SRRM1 as a gene for neurodevelopmental disorder

33 of 37

Case Submission Process Overview

33

34 of 37

Inclusion and Exclusion Criteria

Applicants who are more likely to be accepted:

One or more objective findings.
No diagnosis despite evaluation by specialists who assessed the patient for the objective finding(s).
Agreement with storage and sharing of data and biomaterials .

Applicants who are unlikely to be accepted:

Symptoms with no objective findings.
A diagnosis explaining objective findings.
A diagnosis suggested on record review.
Unwillingness to share data.

34

35 of 37

General Workflow Overview

35

Submit case to the centralized case submission portal

Site gathers records required for case review

Site makes decision and informs participant and their referring provider

Personalized UDN evaluation takes place

(within 60 days of receiving all requested information)

36 of 37

Contact Information

Mount Sinai Center for Undiagnosed Diseases

E: MountSinaiUDN@mssm.edu

T: 646-734-9994

Rachel Evard, MS, CGC (Site Coordinator)

E: rachel.evard@mssm.edu

36

REDCap Inquiry

Center for Undiagnosed Diseases Website

37 of 37

Thank You