Integrated Multi-Omics Analysis
This repository contains notebooks for integrating and analyzing multi-omics data, specifically focusing on RNA-seq, ATAC-seq, Perturb-CITE-Seq, and spatial transcriptomics datasets. The integration is performed using advanced computational methods to provide a comprehensive view of gene expression, chromatin accessibility, and spatial localization. The key steps and methods used across the notebooks are:
-
Melanoma Study using Perturb-CITE-Seq
- A combined therapeutic approach targeting COL1A1 and EIF4A for melanoma treatment.
- A multi-targeted strategy for aggressive melanoma, combining PRAME-targeted immunotherapy with PARP inhibitors and EMT/stemness pathway inhibitors.
- The complex interplay between CD274 (PD-L1), CD58, and CD9 in different immune microenvironments.
- The potential role of LY96 in immune evasion across multiple cancer types, including melanoma.
-
CRISPR Pooled imaging Screen Analysis
-
Over 40 custom geometrical features
-
Vision transformers (DINO) for image feature extraction
-
CellProfiler features
Used these methods to map and cluster genes based on their roles in:
- Protein Synthesis and Quality Control
- Cytoskeleton and Cell Structure
- Intracellular Transport and Organelle Function
-
-
RNA-ATAC Integration with scVI
- Integrated RNA-seq and ATAC-seq datasets using scVI to analyze gene expression and chromatin accessibility data.
- Leveraged publicly available datasets for a comprehensive analysis.
-
RNA-ATAC Integration with FGW and Random Forest
- Utilized the Fused Gromov-Wasserstein (FGW) algorithm to align RNA-seq and ATAC-seq data.
- Quantified the integration using FOSCTTM, Jensen-Shannon Divergence, and cross-entropy metrics.
- Used a Random Forest model to transfer labels between datasets.
-
in silico gene perturbation analysis using the CellOracle package
- Performed pseudotime analysis to order cells along a developmental trajectory
- Analyzed the effects of gene knockouts on downstream gene expression using gene regulatory networks
-
Spatial Mapping using Tangram Package
- Mapped RNA-seq data onto spatial coordinates using a stereo-seq dataset with the Tangram package.
- Analyzed cell-cell interactions at different developmental stages, enhancing the understanding of spatial gene expression patterns.
- Integration of RNA-seq with Spatial Transcriptomics
- Integrated RNA-seq data with spatial transcriptomics (stereo-seq) to map gene expression onto spatial coordinates.
- Facilitated label transfer, analysis of cell-cell communication through ligand-receptor interactions, and investigation of cell-cell transitions using manually curated zebrafish genes for proliferation and apoptosis.
-
Differential Expression Analysis of Immune-Related Genes
- Conducted differential expression analysis of immune-related genes in zebrafish across developmental stages.
- Examined genes related to viral diseases, bacterial infections, and stress responses.