Current Work

Research Projects

A project hub for the lab's integrated work in embryo multi-omics, developmental arrest, and fertility-focused computational biology.

The Digital Embryo

The lab's central program is a multi-omic atlas and predictive model of human preimplantation development, spanning zygote to blastocyst. The project integrates public and newly generated single-cell datasets to identify molecular trajectories, divergence points, and regulatory programs linked to embryo competence and arrest.

Student projects are designed as modular contributions to this shared backbone: transcriptome harmonization, methylation integration, allelic and variant analysis, RNA velocity, time-lapse morphokinetics, and wet-lab validation.

52
Curated datasets in the broader embryo multi-omic landscape
0-6
Developmental days modeled from zygote to blastocyst
4
Program pillars: atlas, dynamics, regulation, validation
1
Shared computational backbone for lab-wide projects
01

Integrated Multi-Omic Atlas of Human Preimplantation Development

Matthew Shen, Jaylan Tran, Dominick Mueller, Arwa Sheheryar, Arsh Verma, Eitan Lenga

This program builds the core Digital Embryo atlas by reprocessing embryo datasets through standardized pipelines, harmonizing metadata, integrating modalities, and comparing normal developmental trajectories against embryo arrest. The goal is to create a durable computational reference that supports downstream prediction, hypothesis generation, and wet-lab validation.

Major Questions
Where do embryos diverge from expected developmental trajectories? Which molecular programs predict progression, arrest, and lineage specification? How can transcriptomic, epigenomic, and time-lapse features be integrated into one interpretable model?
Methods
Single-cell RNA-seqMetadata harmonizationTrajectory inferenceMulti-omic integrationHPC workflows
Related Publication

Russell SJ, Zhao C et al. An atlas of small non-coding RNAs in human preimplantation development. Nature Communications, 2024.

Digital Embryo multi-omic atlas schematic
02

Wet-Lab Multi-Omic Profiling of Human Embryos

Sanaa Ebrahim

This project establishes wet-lab workflows for single-cell and multi-omic profiling of human preimplantation embryos through the CReATe collaboration. The work links sequencing protocol development with the computational atlas so newly generated samples can directly test predictions from the Digital Embryo.

Major Questions
Which protocols best preserve transcriptomic and epigenomic information from limited embryo material? How can new clinical samples validate atlas-derived predictions? What molecular signatures distinguish developmental competence from arrest?
Methods
SS3xpressscNMT-seqEmbryology collaborationSingle-cell multi-omics
Sanaa Ebrahim embryo multi-omic sequencing workflow schematic
03

Methylation and Cross-Modal Integration

Dominick Mueller

This project integrates DNA methylation, transcriptomic, and epigenomic features into a shared model of early development. The immediate focus is to build interpretable cross-modal workflows that let methylation dynamics inform embryo trajectory inference and regulatory architecture.

Major Questions
How do methylation and chromatin features align with transcriptional state transitions? Can cross-modal integration reveal regulatory structure not visible from RNA alone? Which integration strategies are robust across embryo datasets and assay platforms?
Methods
DNA methylationGLUE-style integrationEpigenomicsLatent-space modeling
RNA methylation and epigenome integration workflow for the Digital Embryo
04

Transcriptional Arrest Status Across the Digital Embryo

Arwa Sheheryar, with Jaylan Tran

This project adapts transcriptional arrest status scoring to archival embryo datasets where parental genomes are unavailable. By projecting arrest-related signatures across the atlas, the work tests whether maternal-to-zygotic transition failure can be detected before, during, and after visible developmental arrest.

Major Questions
Can transcriptional arrest be scored without parental whole-genome sequencing? Are arrested cohorts enriched for high-arrest molecular signatures? Do DPRX, ARGFX, HDAC1, and SIRT1-associated signals generalize beyond the original discovery cohort?
Methods
TAS scoringAllelic analysisAUCellscANVIMZT biology
Transcriptional arrest status scoring workflow across embryo datasets
05

Transposable Elements and Alternative Splicing in Embryo Arrest

Arsh Verma

This project quantifies transposable element activity and alternative splicing across developing and arrested embryos. The work asks whether TE families, loci, or splice events mark specific developmental windows or failure modes within the Digital Embryo atlas.

Major Questions
Which TE subfamilies and loci change during progression versus arrest? Are splicing events associated with stage transitions or developmental failure? Can TE and splicing features improve arrest classification?
Methods
SoloTESalmonSUPPA2Splicing analysisFeature matrices
Transposable element and splicing analysis schematic
06

Time-Lapse Imaging and Embryo Morphokinetics

Deeksha Kumar

This project analyzes embryo time-lapse imaging data to connect morphokinetic behavior with developmental potential. The work supports the imaging arm of the Digital Embryo by transforming visual development into quantitative features that can be integrated with molecular data.

Major Questions
Which morphokinetic events distinguish embryos that progress from those that arrest? Can imaging-derived features complement molecular predictors? How should visual developmental dynamics be represented for multi-modal modeling?
Methods
Time-lapse imagingMorphokineticsFeature extractionEmbryo annotation
Embryo time-lapse imaging and morphokinetics schematic
07

Genetic Variant and Allelic Analysis Backbone

Jaylan Tran

This project builds the variant-calling and allelic-analysis layer needed to support embryo-level comparisons across datasets. The work provides a foundation for ancestry-aware quality control, genetic feature annotation, and arrest-related allelic hypotheses.

Major Questions
Can embryo-level variant information be recovered robustly from single-cell RNA-seq? How do genetic and allelic features support cross-dataset quality control? Which variant and allelic signals are useful for arrest-focused analyses?
Methods
MonopogenGATKVCF workflowsPLINKAllelic counts
Digital Embryo variant and atlas analysis visualization
08

RNA Velocity and Developmental Directionality

Eitan Lenga

This project uses RNA velocity to infer developmental directionality from spliced and unspliced transcript counts. By comparing typical trajectories with arrested embryos, the work identifies where transcriptional progression begins to diverge.

Major Questions
How do velocity directionality and confidence differ between progression and arrest? Which genes carry the strongest dynamic signal during early development? Can velocity-derived features improve cross-dataset comparison of embryo trajectories?
Methods
VelocytoscVeloScanpyUMAPTrajectory analysis
RNA velocity UMAP visualization of embryo development