FOG Boston 2026 Agenda

This intensive workshop will provide an overview of multiple approaches to analyze multi-omics data, focusing on tradeoffs between simple and complex approaches to process, integrate, visualise and interpret omics data. In this hands-on session, participants will gain familiarity with different software tools for data processing, analysis and methods to visualise data for model fit interpretation.  

Key Learnings: 

• Gain insight on how multi-omics data can improve our understanding of health and disease 

• Identify study design, data integration and technological gaps and challenges to the use of multi-omics technology and its application to observational studies 

• Define opportunities to overcome these challenges 

Target Audience: 

This workshop is aimed at post-docs, researchers, clinicians, and other professionals with an interest in understanding how omics data can be integrated in clinical and non-clinical research. 

Pre-Requisite Requirements: 

• Introductory background in statistics and data science 

• Familiarity with R 

• Personal laptop and a free, basic RStudio account 

Session 1: Hands-On Multi-Omics: Integrating scRNA-seq and scATAC-seq with Deep Learning for Gene Regulatory Discovery 

• Integrate and analyze paired scRNA-seq and scATAC-seq datasets using 
  modern computational approaches to identify cell types and link 
  chromatin accessibility to gene expression 

• Apply deep learning methods for TF footprinting and sequence 
  attribution to predict transcription factor binding sites and decode 
  regulatory grammar from chromatin accessibility 

• Build reproducible analysis workflows that connect multi-omic 
  measurements to mechanistic gene regulatory insights 

Neva Cherniavsky Durand, Senior Scientist, Broad Institute of MIT and Harvard 

Session 2: Functional Prioritisation of Noncoding Enhancer Variants Using eRNA Regulatory Output and AI-Based Validation 

• Construct reproducible analysis workflows that bridge noncoding variants with post-transcriptional regulation to support mechanistic interpretation of disease-associated variants 
• Quantify how variant-associated changes in eRNA propagate to target gene expression using the TranCi module within eRNAkit at cohort and single-patient resolution 
• Validate predicted regulatory impact by modelling upstream transcription factor (TF) binding disruption and chromatin effects using AI-based sequence tools (DeepSEA) and regulatory annotation platforms (RegulomeDB) 

The session will also emphasise biological interpretation and robustness: participants will run Gene Ontology enrichment analysis, generate interaction plots, and assess how parameter and threshold choices affect prioritisation results and downstream conclusions. 

Natalia Benova, PhD Candidate, Leeds Beckett University