Toma M. Marinov

This paper develops computational methods to predict epitope overlaps in antibodies, showing improved accuracy using contrastive learning and demonstrating their utility in antibody discovery.

This paper introduces MAGE, a model that generates paired human antibody sequences against specific targets, validated against SARS-CoV-2, H5N1, and RSV-A.

The paper maps antibody sequences to antigen specificity across multiple viral antigens, identifying convergent sequence features in antiviral B cells.

This paper designs antigen-specific antibody CDRH3 sequences using AI and germline-based templates, validated against SARS-CoV-2.

This paper introduces a model of reflexive attention that simulates neural attractor states and their role in visual processing, linking them to behavioral and neural correlates like accuracy and EEG components.