Iman Hajirasouliha

A spatial long-read approach at near-single-cell resolution reveals developmental regulation of splicing and polyadenylation sites in distinct cortical layers and cell types.

The paper introduces a spatial long-read sequencing method to study developmental splicing and polyadenylation changes in cortical layers and cell types, finding layer-specific and age-related regulatory patterns.

Spatial isoform sequencing at sub-micrometer single-cell resolution reveals novel patterns of spatial isoform variability in brain cell types

The paper develops a high-resolution spatial long-read sequencing method and software to analyze isoform variability in brain cell types, revealing both region-specific and cell-type-specific patterns.

Blackbird: structural variant detection using synthetic and low-coverage long-reads

Blackbird improves structural variant detection using synthetic and low-coverage long-reads, achieving comparable performance to high-coverage tools with less data.

Automatic Ploidy Prediction and Quality Assessment of Human Blastocyst Using Time-Lapse Imaging

This paper develops BELA, a deep learning model for ploidy prediction in human blastocysts using time-lapse imaging, achieving an AUC of 0.76 on the Weill Cornell dataset.

cloudrnaSPAdes: Isoform assembly using bulk barcoded RNA sequencing data

cloudrnaSPAdes assembles full-length isoforms from barcoded RNA-seq linked-read data, accurately handling high isoform diversity in simulated and real human data.

Ariadne: Synthetic Long Read Deconvolution Using Assembly Graphs

Ariadne deconvolves synthetic long read data using assembly graphs to improve taxonomic classification and de novo assembly of complex populations.

CAMP: A modular metagenomics analysis system for integrated multi-step data exploration

CAMP provides a modular metagenomics analysis system that enables scalable, extensible workflows with integrated data visualization and reporting.

Weakly-Supervised Tumor Purity Prediction FromFrozen H&E Stained Slides

This paper introduces a weakly-supervised method to predict tumor purity from H&E stained slides, achieving high accuracy and enabling clinical applications.

A deep learning framework for structural variant discovery and genotyping

This paper introduces a deep learning framework called Cue for structural variant discovery and genotyping, showing improved performance over existing methods on synthetic and real data.

coronaSPAdes: from biosynthetic gene clusters to RNA viral assemblies

coronaSPAdes improves RNA viral genome assembly, particularly for coronaviruses, by leveraging viral genome structure knowledge.

Shotgun Transcriptome and Isothermal Profiling of SARS-CoV-2 Infection Reveals Unique Host Responses, Viral Diversification, and Drug Interactions

This study uses rapid diagnostics and metatranscriptomics to analyze SARS-CoV-2 infection, revealing viral evolution, host responses, and drug interactions in NYC samples.

Inferring Cancer Progression from Single-cell Sequencing while Allowing Mutation Losses

The paper introduces SASC, a method for inferring cancer progression from single-cell sequencing data that allows mutation losses, showing high accuracy on simulated and real datasets.

Characterization of segmental duplications and large inversions using Linked-Reads

The paper introduces VALOR2, a new software package that uses Linked-Read sequencing to detect large genomic inversions and characterize interspersed segmental duplications.

PhISCS - A Combinatorial Approach for Sub-perfect Tumor Phylogeny Reconstruction via Integrative use of Single Cell and Bulk Sequencing Data

PhISCS integrates single-cell and bulk sequencing data to reconstruct sub-perfect tumor phylogenies by minimizing false negatives, false positives, and ISA violations.

A robust benchmark for germline structural variant detection

Genome in a Bottle benchmark of 12745 sequence-resolved insertions and deletions of 50 bp or more, integrating 19 calling methods, scores false negatives and false positives alike.

Detection and assembly of novel sequence insertions using Linked-Read technology

The paper introduces Novel-X, a method using Linked-Read data to detect novel sequence insertions without whole-genome de novo assembly, outperforming short-read tools and enabling cost-effective large-scale screening.

Novel Algorithms for the Taxonomic Classification of Metagenomic Linked Reads

KrakenLinked improves taxonomic classification of metagenomic linked-reads by combining tree pruning and taxa promotion algorithms, reducing false identifications compared to short reads.

Minerva: An Alignment and Reference Free Approach to Deconvolve Linked-Reads for Metagenomics

Minerva deconvolves barcoded reads in metagenomics to improve taxonomic assignments and sequence clustering without a reference genome.

Robust Automated Assessment of Human Blastocyst Quality using Deep Learning

This paper develops an AI system to assess human blastocyst quality using deep learning, achieving 98% accuracy and identifying clinical factors affecting pregnancy rates.

Breast Cancer Histopathological Image Classification: A Deep Learning Approach

Fine-tuned ImageNet ResNets beat Inception models on tissue micro-array and BreaKHis slides, reaching 99.8% across four tumour tissues and 96.4% across four malignant sub-types.

Deep Convolutional Neural Networks Enable Discrimination of Heterogeneous Digital Pathology Images

This study develops a CNN-based pipeline to classify histopathology images across cancer types and biomarkers, achieving high accuracy in distinguishing cancer subtypes and IHC scores.