Marek Mutwil

The paper constructs a kingdom-wide stress transcriptome atlas of 36 Viridiplantae species, revealing conserved stress response programs and a growth-defence tradeoff across 700 million years of divergence.

This paper introduces a biological spike-in method for multi-species RNA-seq, revealing transcriptional collapse under dark stress in Arabidopsis and Brachypodium.

This paper creates a structured atlas of research logic in Plant Cell by converting articles into Research Process Graphs using LLMs, revealing patterns in research methods and findings.

ORBIT aligns cross-species coexpression network embeddings using orthogonal Procrustes rotation, achieving higher cross-species retrieval quality than autoencoders.

This paper introduces TEA-GCN, a method for constructing gene co-functional and co-regulatory networks from public RNA-seq data, showing improved performance and cross-species conservation.

This paper evaluates LLMs for plant biology tasks, identifies their limitations, and proposes GraphRAG as a solution using structured knowledge graphs for reliable, reproducible answers.

This study identifies conserved and lineage-specific genes in pollen and anther development using cross-species RNA-seq analysis and experimental validation.

The paper compares abiotic stress response networks across three plant species, finding conserved gene sets and functional divergence in transcription factors.

PlantConnectome extracts 5 million functional relationships from 71,000 plant biology abstracts using large language models and provides a database for gene-metabolite interactions and regulatory networks.

This study identifies NYH0002 as a compound that blocks ILF2, leading to DNA breaks and lethality in cancers with specific genetic profiles.

The paper generates transcriptomic data for 22 fern species to study their evolutionary genetics and cell wall composition, revealing unique gene families and divergent cell wall biochemistry.

The paper introduces a yeast knowledge graph database built using AI to extract gene function relationships from publications, providing a searchable resource for yeast research.

The paper extracts and analyzes connectivity information from publications to build searchable connectomes for C. elegans and Drosophila using AI and bioinformatics tools.

The paper demonstrates that disrupting super-silencers via EZH2 and REST inhibition reduces chromatin interactions and cancer growth.

The paper updates the diurnal.plant.tools database by adding diurnal gene expression data from four crop species and Marchantia polymorpha, enhancing cross-species comparisons and evolutionary insights into circadian clocks.

LSTrAP-denovo generates transcriptome atlases for eukaryotic species without genomes by assembling and annotating RNA-seq data.

The paper uses k-means clustering to identify condition-specific co-expressed genes in tea plants, finding a gene involved in the bud-prealbinism stage of 'Anji Baicha'.

The paper constructs a comparative transcriptome database for Camellia sinensis to identify genes involved in flavonoid synthesis across tea cultivars.

The paper constructs a gene feature dataset for Arabidopsis thaliana and develops a machine learning workflow to identify linked features, visualized as a Feature Important Network (FIN) to reveal novel biological insights.

The paper identifies a conserved genetic program in male germ cells across species, revealing 161 new spermatogenesis genes and three potential causes of human infertility.

Bacteria.guru provides a database for comparative transcriptomics and co-expression networks of bacterial pathogens to aid in understanding gene function and antibiotic resistance.

The paper computationally analyzes RNA secondary structures of viral genomes from multiple hemorrhagic fever species, finding structural differences linked to virulence and human protein interactions.

The paper develops a negative selection method to remove auto-activators from yeast-two-hybrid assays using pGAL2-URA3 and 5-FOA treatment, reducing false-positive hits.

LSTrAP-Kingdom automates gene expression atlas generation across kingdoms by processing RNA-seq data and producing annotated matrices comparable to manual curation.

The paper uses crowdsourced analysis of RNA-seq data to identify uncharacterized genes associated with protein synthesis in bacterial pathogens.

The paper introduces fungi.guru, a database integrating genomic and transcriptomic data from 19 fungi to identify conserved pathways and gene functions, with examples of secondary metabolite and catabolic pathways.

LSTrAP-Cloud is a cloud-based pipeline that infers gene co-expression networks from RNA sequencing data to identify functionally related genes.

The paper demonstrates using a low-cost computer to predict gene function and infer regulatory networks from RNA sequencing data in Artemisia annua.

This paper presents the avocado genome, revealing its phylogenetic position, introgressive hybridization, and pathogen-driven gene space adaptation.

The paper introduces Malaria.tools, a database that integrates genomic and transcriptomic data from Plasmodium species to predict gene function and identify pathogenicity-related genes.