Related Experiment Video
Updated: Oct 10, 2026
Generating the Transcriptional Regulation View of Transcriptomic Features for Prediction Task and Dark Biomarker Detection on Small Datasets
Published on: March 1, 2024
Transcriptomic buffering detection: a computational approach to identifying essential gene dependencies in pancreatic
Magnus Boman1, Eleni Afentaki2, Julia Marie Greely3
1Division of Clinical Epidemiology, Department of Medicine Solna, Karolinska Institutet, SE-171 76 Stockholm, Sweden.
Motivation:
Loss-of-function (LOF) mutations in essential genes (Goners) should be lethal, yet some tumours survive through compensatory buffering mechanisms. Existing computational methods for identifying synthetic lethal (SL) interactions focus on known tumour suppressor losses, missing essential gene dependencies. We developed a computational pipeline to identify SL buffering mechanisms by correlating Goner LOF burden with transcriptomic compensation.
Results:
Applying our pipeline to pancreatic adenocarcinoma (PAAD), we integrated CRISPR essentiality screens (DepMap), mutation data, and RNA expression profiles from 165 TCGA-PAAD cases. This PAAD analysis supports the feasibility of the pipeline but does not nominate a transcriptomic buffer; applicability to other cancer types remains to be tested.
Availability And Implementation:
Analysis code and documentation are available at https://github.com/d3dd/goblet. Implemented in Python 3.9+ using pandas, scipy, and gseapy. All data are publicly available from TCGA and DepMap.