Alex Zelikovsky

The paper predicts gene mutations in colon cancer using a gene-conditioned LSTM model with attention on inferred mutation paths from a directed co-occurrence asymmetry graph.

The paper benchmarks discrepancies across microbial genome reference databases, finding high consistency in viral genomes but significant variability in fungal genomes.

FASTR is a new format for storing sequencing data that reduces file size and improves processing speed compared to FASTQ.

This paper introduces a hybrid training model to enhance bioinformatics capacity globally, demonstrating its effectiveness through an Eastern European workshop series.

This study benchmarks 12 HLA callers on RNA-seq data across 682 samples to evaluate accuracy, computational cost, and performance differences by ancestry.

The paper develops a computational framework to detect emerging viral variants with altered phenotypes by analyzing epistatic networks, achieving early detection of known SARS-CoV-2 strains.

This paper clusters SARS-CoV-2 sequences to identify variants and subtypes, showing improved performance over existing methods and detecting rare variants in large datasets.

This paper introduces CliqueSNV, a method to accurately identify minority viral haplotypes from NGS data by reducing sequencing noise, achieving detection at frequencies as low as 0.1%.

The paper compares intra-host HCV population characteristics between recent and persistent infections, finding significant differences and developing a machine learning classifier with 95.22% accuracy.

The paper normalizes heterogeneous genomic data into images and applies machine learning to classify viral infection stages and detect transmission clusters with high accuracy.

This paper introduces a method using k-mer distribution in RNA-seq reads to estimate genetic relatedness between viral samples and infer transmission clusters without requiring full RNA-seq assembly.

This study benchmarks error-correction methods for next-generation sequencing data, finding significant variation in performance across datasets with no single method being optimal overall.

This paper introduces SCIFIL, a tool to infer fitness landscapes of cancer clones from single-cell sequencing data, demonstrating its accuracy on simulated and experimental data.

The paper develops ImReP, a computational method to profile TCR and BCR repertoires from RNA-seq data, creating a large atlas of immunological sequences across 53 tissues.

The paper presents a fast and scalable k-mer-based framework for efficiently extracting genetically similar sequences from large NGS datasets, targeting heterogeneous populations like viruses.

This paper introduces BATISCAF, a repeat-aware scaffolder that outperforms existing tools on benchmark datasets.

This paper evaluates scaffolding tools by accounting for repeats in genomic sequences, proposing a repeat-aware evaluation framework and applying it to assess state-of-the-art tools on GAGE datasets.

This paper introduces 2SNV, a method to accurately reconstruct rare viral mutant variants from single-molecule sequencing data with high sensitivity and accuracy.

HapIso reconstructs haplotype-specific isoforms from long single-molecule reads by clustering reads into parental alleles despite high error rates.