Perturbation-aware representation learning for heterogeneous single-cell genetic screens
Manuscript · In preparation
Computational biology · machine learning
Tanmay Tanna
I develop statistical and machine-learning methods for complex data and lead their application in large-scale biological studies.
At ETH Zürich, my work spans genome-scale in vivo Perturb-seq, CRISPR-based transcriptional recording, metabolomics and cancer multi-omics, from experimental design and scalable analysis to reusable software.
Research focus
Methods grounded in biological questions
I apply statistical methodology and machine learning across experimental domains: cancer biology, genetic perturbations and transcriptional recording, with an emphasis on approaches that remain useful beyond a single dataset.
Perturbation modelling and in vivo screens
Designing genome-scale perturbation experiments and developing statistical and machine-learning methods to interpret their effects in live mice.
Read moreTranscriptional recording
Computational methods for CRISPR-based systems that turn bacteria into recorders of their own molecular activity.
Read moreMulti-omics method development
Statistical and algorithmic methods for metabolomics, metagenomics and multimodal data.
Read moreComputational cancer research
Multimodal analysis of clinically annotated cancer cohorts, with work in melanoma and ovarian cancer.
Read moreSelected outputs
Recent and foundational work
Principal contributions to research spanning transcriptional recording, single-cell perturbation screens, cancer multi-omics and computational method development.
An end-to-end computational framework for ‘Record-seq’ transcriptional recording data
Bioinformatics · Accepted
Recording transcriptional histories using Record-seq
Nature Protocols
Contact