Sina Majidian

Omni2Tree directly analyzes raw sequencing reads to reconstruct viral phylogenies and identify diversification, outperforming assembly and reference-based methods in low-coverage scenarios.

ARCLID is a deep learning-based structural variant caller that achieves high accuracy at low sequencing coverages, outperforming existing tools and enabling cost-effective SV discovery.

Movi Color improves long-read classification accuracy and speed using a move structure and color-based indexing, outperforming existing methods in precision and recall.

The paper benchmarks ANI estimation algorithms using simulated and real datasets, showing that ANIb is most accurate but least efficient, while k-mer approaches are efficient and accurate, with some clades requiring multiple k-values.

The paper introduces PoMoCNV, a likelihood-based method to infer evolutionary parameters of CNVs using population genetics data, validated on C. elegans datasets.

FastOMA scales orthology inference to handle large genome datasets efficiently while maintaining accuracy.

The paper introduces genomic stratifications for human reference genomes GRCh37, GRCh38, and T2T-CHM13v2.0 to benchmark variant caller performance in difficult regions.

Assembling the Sequence Read Archive (27.3 million accessions) with over 100-fold compression enables minutes-scale k-mer search and mining of enzymes, plasmids, proteins and viral reactivation.

Read2Tree processes raw sequencing reads to build phylogenetic trees, outperforming conventional methods in speed and accuracy across diverse datasets.

OMAMO selects the best simple model organism for studying a biological process based on orthologous relationships between humans and species.

Hap10 reconstructs accurate and long polyploid haplotypes using linked reads, validated on sweet potato data.