Author

Debjani Das

Victor Chang Cardiac Research InstituteORCID 0009-0008-5654-7079Unclaimed ORCID profile · follow to get notified of new preprints

Additional preprints under this name without ORCID

Debjani Das2 papers

Biology Disciplines

Click to filter author papers

Bioinformatics· 1 Genomics· 1

Signature Assays, Organisms & Methods

Click to filter · or open tag

Research Focus:benchmarking· 1 Organism:human· 1 Approach:large language model· 1 Approach:statistical model· 1 Research Focus:variant effect prediction· 1 Assay:whole-exome sequencing· 1 Assay:whole-genome sequencing· 1

Frequent Co-Authors

Eleni GiannoulatouiD· 2 shared Sally L. DunwoodieiD· 2 shared Congenital Heart Disease Synergy Study Group· 1 shared David Scott WinlawiD· 1 shared Edwin Philip KirkiD· 1 shared Gary F. SholleriD· 1 shared Gavin ChapmaniD· 1 shared Gillian M. BlueiD· 1 shared

Activity by Year

2024-2026 · peak 1/yr

Co-Author Collaboration Network

10 top collaborators · 38 mutual co-author links

Click any collaborator node to inspect or filter shared papers

Focal author (center)Selected co-authorORCID co-authorUnlinked name

Curved chords = mutual co-authorship

2 shared papersBioinformatics

Eleni GiannoulatouORCID

Mutual co-authors in cluster

Author profile →

2 papers

Recent·Most cited

BioinformaticsbioRxivPreprintSep 15, 2026

HeartVar: An LLM-Assisted Tool for Clinical Classification of Variants in Cardiovascular Disease Cohorts

Jamie‐Lee M. Thompson, Debjani Das, Sally L. Dunwoodie, Eleni Giannoulatou

HeartVar automates variant classification for cardiovascular disease using a hybrid rule-based and LLM approach, achieving 72% accuracy on benchmarked ClinGen variants.

DOI 10.64898/2026.09.10.750569

humanwhole-exome sequencing

𝕏 PostBluesky

GenomicsBioinformatics AdvancesPeer-reviewedDec 26, 2024

Benchmarking of variant pathogenicity prediction methods using a population genetics approach

Mikhail Gudkov, Loïc Thibaut, Steven Monger…·Samantha J. Lain

This study benchmarks variant pathogenicity predictors using population genetics data, finding CADD and REVEL as the best performers and highlighting biases in ClinVar-based training.

3 citations

whole-genome sequencingstatistical model

𝕏 PostBluesky