Maria O. Anisimova

The paper identifies STR variation in olive and evaluates how genomic language models respond to natural STR length changes, finding consistent embedding perturbations and prioritizing candidate loci linked to oil composition.

This paper introduces a framework using STR variations for high-resolution population structure inference, showing STRs provide finer resolution than SNPs.

ConSTRain is a novel STR variant caller that accounts for copy number alterations, enabling accurate genotyping in polyploid and aneuploid organisms.

The paper evaluates codon optimization approaches for protein expression in E. coli, finding that negative epistasis limits prediction accuracy and that non-proprietary methods can achieve high expression.

The paper identifies STR variations associated with gene expression and DNA methylation across three cancer types and finds shared regulatory mechanisms.

The paper evaluates ARPIP's ability to reconstruct ancestral sequences with long indels, showing it preserves gap length distributions and captures deletion bias in mammalian proteins.

This paper introduces indelMaP, a method that improves ancestral sequence reconstruction and multiple sequence alignment by treating insertions and deletions as distinct evolutionary events, achieving high accuracy and speed for densely sampled datasets.

This study identifies 1244 expression STRs (eSTRs) in colorectal cancer, showing their association with gene expression and mutability in MSI tumors, suggesting potential therapeutic targets.

The paper identifies a novel hyper-VNTR in SLC6A3, characterized by high heterozygosity, G-quadruplex formation, and potential recombination hotspot properties.

The paper introduces a multi-channel deep neural network for biomedical relation extraction that outperforms existing methods by leveraging textual and graph features.

This paper introduces a combinatorial optimization method to jointly infer multiple sequence alignments and phylogenetic trees under maximum likelihood, addressing computational limitations of existing approaches.

The paper introduces a semantic model and federated approach to integrate heterogeneous biological databases, enabling joint queries across gene expression, orthology, and protein data sources.

The paper investigates the impact of -1 programmed ribosomal frameshifting on codon usage bias indices and finds that corrections for this effect strengthen the translational efficiency hypothesis.