Manu Saraswat

Mapping trans-eQTLs at single-cell resolution using Latent Interaction Variational Inference.

This paper introduces LIVI, a deep learning framework for mapping trans-eQTLs at single-cell resolution, and shows it detects more trans-eQTLs than existing methods and identifies cell types and genes involved in autoimmune disease risk.

High-throughput single-cell CRISPRi screens stratify neurodevelopmental functions of schizophrenia-associated genes

The paper uses single-cell CRISPRi screens to functionally assess schizophrenia-associated genes, revealing disruptions in neurodevelopmental timing and identifying key regulatory factors.

A spatiotemporal cancer cell trajectory underlies glioblastoma heterogeneity

The paper identifies a spatiotemporal cancer cell trajectory in glioblastoma that explains tumor heterogeneity across regions and subclones.

Decoding Plasticity Regulators and Transition Trajectories in Glioblastoma with Single-cell Multiomics

This study uses single-cell multiomics to identify regulatory programs and plasticity regulators in glioblastoma, revealing MYT1L as a key repressor of malignant plasticity.

ExplaiNN: interpretable and transparent neural networks for genomics

ExplaiNN provides interpretable deep learning models for genomics tasks, achieving comparable performance to state-of-the-art methods while offering transparent predictions.

Biologically-relevant transfer learning improves transcription factor binding prediction

This study shows that transfer learning improves transcription factor binding prediction by pre-training with biologically relevant TFs and fine-tuning single-task models.

Discovery of widespread transcription initiation at microsatellites predictable by sequence-based deep neural network.

The paper identifies widespread transcription initiation at microsatellites across species and develops deep learning models to predict STR-associated transcription and its impact on disease variants.

Evidence of transcription at polyT short tandem repeats

The paper investigates transcription initiation at polyT short tandem repeats (STRs) using CAGE data and develops a deep learning model to predict STR-associated transcription with high accuracy.