Andrei Slabodkin

TL;DR: In this paper, the authors identify structural interaction motifs, which together compose a commonly shared structure-based vocabulary of paratope-epitope interactions, and show that this vocabulary enables the machine learnability of antibody-antigen binding using generative machine learning.

TL;DR: The largest available set of non-redundant antibody-antigen structures are screened for binding patterns and identified structural interaction motifs are identified, which together compose a vocabulary of paratope-epitope interactions that is universally shared among investigated antibody-antsigen structures.

TL;DR: The Absolut! as mentioned in this paper software suite enables the generation of synthetic lattice-based 3D-antibody-antigen binding structures with ground-truth access to conformational paratope, epitope, and affinity.

TL;DR: In this article, the authors present an open-source collaborative ecosystem for machine learning analysis of adaptive immune receptor repertoires (AIRR) and demonstrate the broad applicability of immuneML by reproducing a large-scale study on immune state prediction, developing, integrating and applying a novel deep learning method for antigen specificity prediction and showcasing streamlined interpretability-focused benchmarking of AIRR ML.

TL;DR: In this article, a sensitivity-tested distance measure was devised to enable inter-individual comparison of VDJ recombination models, and it was shown that population-wide individualized recombination can result in orders of magnitude of difference in the probability to generate (auto)antigen-specific Ig sequences.