Protein foundation models increasingly support the generation of novel biological sequences, yet their reliance on large public datasets creates opportunities for adversarial manipulation of the material they learn from and retrieve. We investigate two data-only poisoning attacks against protein design systems, targeti...
Charalampos Koilakos, K. Provatas, Michail Patsakis et al.· 0 citations
BarcodesDB is presented, a comprehensive database of synthetic DNA barcodes systematically identified by scanning 403 199 complete organismal genome assemblies across the tree of life, together with 215 Gbp of raw metagenomic sequencing reads spanning marine, soil, polar and host-associated environments, providing maxi...
Michail Patsakis, K. Provatas, I. Mouratidis et al.· Bioinformatics Advances· 0 citations
Abstract Motivation Genomic and protein foundation models (GFMs and PFMs) have demonstrated strong performance in learning the language of DNA and proteins, but their use in large-scale sequence generation is limited by the latency of autoregressive decoding. Because every token triggers a forward pass of a large Trans...
K. Provatas, Aris Karatzikos, Charalampos Koilakos et al.· Bioinform.· 0 citations
TurboQuant-Bio, an open toolkit that compresses Evo 2’s weights and attention cache to four bits without calibration data, and serves both through fused kernels, proves near-lossless across perplexity spanning the tree of life, genomic classification, splice-site prediction, gene completion and clinically relevant vari...
Michail Patsakis, Alexandros Tzanakakis, I. Georgakopoulos-Soares· bioRxiv· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
