S. Joshua Swamidass

Standard operating procedure for somatic variant refinement of tumor sequencing data

This paper establishes a standard operating procedure for manual refinement of somatic variants in tumor sequencing data to improve consistency and reproducibility.

Deep Learning Global Glomerulosclerosis in Transplant Kidney Frozen Sections

This paper develops a deep learning model to classify glomeruli as normal or sclerotic in frozen kidney biopsy slides, achieving performance comparable to expert pathologists with limited training data.

Opportunities And Obstacles For Deep Learning In Biology And Medicine

Deep learning rivals or beats baselines across imaging, health records, genomics and drug discovery but is not yet transformative; scarce labels, privacy and opacity remain hurdles.

Evolution and Functional Information

This paper challenges the reliability of 'functional information' as a measure of evolutionary difficulty by showing that sequence alignment mutual information underestimates functional proteins and does not reflect functional proximity.