From data upload to reports, AI connects your entire workflow
Less environment setup and tool switching. More focus on biological questions.
AI-powered analysis, from start to finish
Upload your data
Upload sequencing files, expression matrices or sample metadata.
Ask in natural language
Describe your goals, groups and analysis requirements.
Build the workflow
Create an analysis DAG with tools and parameters selected by AI.
Run the analysis
Execute tools and scripts while retaining intermediate outputs.
Generate the report
Bring together charts, statistics, methods and biological interpretation.
Ready for your research
58 analysis Skills, 46 tools and 36 database/data-source entries, covering data processing through interpretation. Of these data sources, 22 are integrated and 14 are not yet integrated.
Skills
alignment-gap-trimming-metrics
Find raw FASTA files or precomputed ortholog alignments, run MAFFT when needed, calculate gap proportions with explicit denominators and aggregation rules, and evaluate and rank ClipKIT-style trimming reductions for eligible complete orthologs.
archive-dataset-prep
Safely extract or stage archived bioinformatics datasets, preserve group and sample boundaries, identify file roles, and produce a manifest for downstream analysis.
biomedical-literature-db-lookup
Turn retrieved biomedical records and local methods evidence into explicit analysis constraints, reproducible sample exclusion rules, entity mappings, database accessions and handoff artifacts.
biostatistics-tests-power
Identify small biological experiment tables and run pooled-SD Cohen's d, t-tests, Welch and Mann–Whitney U tests, Shapiro–Wilk normality tests, one- or two-way ANOVA, and two-sample power or sample-size analysis with auditable numerical results.
bulk-rnaseq-deseq2-core
Plan and run bulk RNA-seq differential expression with DESeq2 or PyDESeq2 from count matrices and sample metadata, including schema detection, design formulas, contrasts and reference groups, effect-size shrinkage choices and differential-expression thresholds.
busco-ortholog-completeness
Find, parse and summarize BUSCO tables and single-copy sequences; calculate complete single-copy orthologs shared across proteomes, total amino acids and BUSCO status summaries with explicit lineages, sample groups and denominators.
clinical-data-qc-filtering
Check patient- or participant-level clinical tables, including complete-case counts across clinical and omics data, missing or invalid labels, exact AE/MH/DM event filtering, anomalous cohort isolation and identifier alignment.
clinical-ordinal-logistic-vaccine
Identify AE, DM and MH trial tables, construct vaccine-related ordinal severity cohorts, and run proportional-odds logistic regression or chi-square subgroup tests. Report treatment effects, interactions, medical history, odds ratios, confidence intervals, p-values and percentage changes.
clinical-treatment-response-logistic
Fit binary logistic models for clinical treatment response with explicit cohorts and event coding, compare formulas and AIC, and report coefficients, p-values, odds ratios and predicted probabilities.
consensus-clustering-pca-enrichment
Analyze expression-based subtypes through input and schema detection, sample alignment, duplicate handling, consensus or bootstrap clustering, PCA explained variance and per-cluster Reactome, GO, KEGG or WikiPathways enrichment.
correlation-expression-essentiality
Identify table schemas and analyze relationships between gene expression and gene length, essentiality, dependency or expression-distribution skewness using Pearson or Spearman correlations and FDR/BH correction, with auditable gene-level results.
cpg-methylation-density
Identify CpG-site and chromosome-length tables, filter by methylation thresholds, count distinct sites, and calculate chromosome densities, density ratios, the highest-density chromosome, genome-wide mean density, length-weighted chi-square statistics and removed-site counts.
crispr-screen-enrichment
Analyze CRISPR or MAGeCK screens by identifying gene and sgRNA tables, calculating replicate correlations, condition-level pathway overrepresentation, strongest enrichment, stability across rounds, significant-condition proportions, immune pathways and preranked GSEA results.
data-file-inspection
Inspect bioinformatics inputs, directories, table schemas, Excel sheets, sample identifiers and analysis-ready artifacts in read-only mode to establish downstream data availability.
debugging-artifact-contracts
Diagnose failed bioinformatics runs and validate artifact paths, summaries, answer files and final-answer requirements before reporting results.
gene-annotation-id-mapping
Identify and convert local or batch gene identifiers, combine NCBI or Ensembl records, and produce reproducible annotation mappings and downstream handoff artifacts.
human-chip-variant-cohort
Analyze human BLM/CHIP cohort variants, including sample metadata joins, VAF thresholds and bins, CHIP membership, non-reference or exonic filtering, benign/synonymous/missense proportions, per-sample counts, leading genes and group frequency tests.
kegg-enrichment-specialist
Run reproducible enrichment on RNA-seq or differential-expression results using a fixed local KEGG snapshot, Python overrepresentation analysis or offline-adapted clusterProfiler, checking species, schemas, contributing genes and pathway-set relationships.
m6a-expression-integration
Integrate m6A or differential methylation with expression or differential-expression status. Identify schemas and counting units, calculate hyper/hypomethylation counts, proportions, ratios and m6A-by-expression contingency tables, and run Pearson chi-square independence tests.
metabolomics-untargeted-lcms
Process untargeted LC-MS raw data with XCMS peak detection, retention-time correction, feature grouping and missing-peak filling, or analyze feature tables with two-group PCA, Welch or Wilcoxon tests, FDR correction and visualization.
microbial-read-qc-coverage
Interpret microbial paired-end read QC, run or audit Trimmomatic trimming, prepare alignments with BWA-MEM and samtools, and calculate per-sample mean coverage with explicit read/read-pair units and reference denominators.
microbial-variant-calling-summary
Find microbial VCF, BAM and reference files, run or audit GATK HaplotypeCaller as needed, summarize SNP/Indel allele classes, calculate raw or QUAL-filtered transition/transversion ratios and compare mutation accumulation across samples.
mirna-expression-analysis
Analyze case–control qPCR/Ct matrices or immune-cell miRNA expression, including schema detection, sample exclusion, log transformation, Welch tests, multiple-testing correction, miRNA denominators, LFC matrices, ANOVA and median comparisons.
pathway-ora-gsea-general
Identify differential-gene tables, gene lists, ranked lists or enrichment results and run or query ORA/GSEA using Reactome, WikiPathways, GO, KEGG, Enrichr/gseapy or local GMT files. Return significance, odds ratios, overlapping genes and leading-edge genes.
phenotype-assay-summary
Identify phenotype assay tables and calculate descriptive statistics, standard errors, percentage changes, ratios, raw group differences, nearest conditions and Dunnett-style control comparisons with auditable numerical outputs.
phylogenetic-tree-metrics
Find grouped phylogenetic trees and alignments; calculate PhyKIT-compatible treeness, long-branch scores, lineage distances, tree length, DVMC, parsimony-informative sites, RCV, treeness/RCV and evolutionary rates, then summarize values, percentiles, ratios or Mann–Whitney tests.
rds-bioconductor-io
Inspect, read and export RDS/RData bioinformatics objects, including data frames, DESeq2 results, SummarizedExperiment, SingleCellExperiment and Seurat components, with an explicit choice of pyreadr or Rscript.
report-amplicon-16s-its-structured
Generate structured 16S/ITS amplicon reports from newly generated, auditable ASV/OTU, taxonomy, alpha/beta diversity, differential taxa, association and functional-prediction evidence. Report generation does not replace analysis execution.
report-artifact-delivery-navigator
Create artifact-focused delivery guides for newly generated files, figures, code and formal reports, including summaries, deliverable indexes, file manifests and output navigation pages.
report-atac-seq-structured
Generate structured ATAC-seq reports from newly generated, auditable accessibility, peak, differential-accessibility, motif, transcription-factor and regulatory-network evidence. Report generation does not replace analysis execution.
report-bulk-rnaseq-differential-structured
Generate structured bulk RNA-seq differential-expression reports from newly generated, auditable count matrices, sample relationships, differential-expression and enrichment evidence. Report generation does not replace analysis execution.
report-chip-seq-structured
Generate structured ChIP-seq reports from newly generated, auditable protein-binding, peak, differential-binding, motif, co-binding and regulatory-network evidence. Report generation does not replace analysis execution.
report-clinical-decision-brief
Prepare concise clinical decision briefs, multidisciplinary-team briefs, risk-stratification reports or recommendation summaries from newly generated patient, cohort, biomarker, treatment-response or diagnostic results.
report-dna-methylation-structured
Generate structured DNA methylation reports from newly generated, auditable methylation landscapes, differential sites and regions, promoters, enrichment and expression associations. Report generation does not replace analysis execution.
report-executive-decision-summary
Turn newly completed analyses, projects, experiments or operational reviews into management briefs, executive summaries, key-metric dashboards, go/no-go decision memos or leadership action reports.
report-genome-denovo-assembly-structured
Generate structured de novo genome assembly and annotation reports from newly generated, auditable assembly, polishing, scaffolding, completeness, contamination, annotation and comparative-genomics evidence. Report generation does not replace analysis execution.
report-hi-c-structured
Generate structured Hi-C chromatin-conformation reports from newly generated, auditable contact matrices, compartments, TADs, loops, enhancer–promoter and differential-interaction evidence. Report generation does not replace analysis execution.
report-immune-repertoire-structured
Generate structured immune-repertoire reports from newly generated, auditable clonotype, CDR3, VDJ usage, diversity, clonal expansion, differential clonotype and antigen-association evidence. Report generation does not replace analysis execution.
report-metabolomics-structured
Generate structured metabolomics reports from newly generated, auditable metabolite identification, multivariate models, differential metabolites, pathways and biomarker evidence. Report generation does not replace analysis execution.
report-methods-reproducibility-audit
Produce detailed methods and reproducibility audits for newly executed analyses or workflows, including parameter tracking, quality-gate records and rerun checklists.
report-multiomics-integration-structured
Generate structured multi-omics integration reports from newly generated, auditable single-omics results, cross-omics mappings, correlations, joint factors, networks and shared pathways. Report generation does not replace analysis execution.
report-pathogen-mngs-structured
Generate structured pathogen mNGS reports from newly generated, auditable microbial detection, read and coverage support, control backgrounds, contamination, resistance and virulence evidence. Report generation does not replace analysis execution.
report-population-genomics-gwas-structured
Generate structured population-genomics and GWAS reports from newly generated, auditable population structure, diversity, selection, association, fine-mapping and candidate-gene evidence. Report generation does not replace analysis execution.
report-project-analysis-structured
Produce complete project analysis reports from newly generated evidence, covering cover information, reproducibility appendices, executive summaries, project design and workflows, results, discussion and next steps.
report-proteomics-structured
Generate structured proteomics reports from newly generated, auditable peptide/protein identification, quantification, differential proteins, enrichment, interactions and post-translational modifications. Report generation does not replace analysis execution.
report-scientific-imrad-manuscript
Use newly completed research evidence to draft scientific reports or manuscripts in Introduction, Methods, Results and Discussion format, with optional abstracts and other standard sections.
report-shotgun-metagenomics-structured
Generate structured shotgun metagenomics reports from newly generated, auditable taxonomic composition, diversity, functional profiles, resistance, virulence, assemblies and metagenome-assembled genomes. Report generation does not replace analysis execution.
report-single-cell-rnaseq-structured
Generate structured single-cell RNA-seq reports from newly generated, auditable cell QC, clustering, annotation, composition, differential states, trajectories and cell-communication evidence. Report generation does not replace analysis execution.
report-small-rna-seq-structured
Generate structured small RNA-seq reports from newly generated, auditable small-RNA composition, miRNA identification, differential miRNAs, target genes and regulatory networks. Report generation does not replace analysis execution.
report-spatial-transcriptomics-structured
Generate structured spatial-transcriptomics reports from newly generated, auditable image registration, spatial expression, tissue regions, deconvolution and spatial interactions. Report generation does not replace analysis execution.
report-tumor-somatic-variants-structured
Generate structured tumor somatic-variant reports from newly generated, auditable somatic variants, tiering and classification, biomarkers and relevant negative findings. Report generation does not replace analysis execution.
report-wes-wgs-germline-structured
Generate structured WES/WGS germline-variant reports from newly generated, auditable SNVs, Indels, CNVs, structural variants, annotations, phenotypes and family cosegregation evidence. Report generation does not replace analysis execution.
report-whole-transcriptome-ncrna-structured
Generate structured whole-transcriptome noncoding-RNA reports from newly generated, auditable mRNA, lncRNA, circRNA, miRNA, differential transcripts, targets and ceRNA networks. Report generation does not replace analysis execution.
rnaseq-go-enrichment
Run DESeq2 differential expression and GO biological-process enrichment on bulk RNA-seq count matrices, with targeted GO-term queries, leading-term simplification, differential-gene numerical queries, gene-ID mapping and enrichment background selection.
rnaseq-result-query-overlap
Identify existing RNA-seq result schemas, query gene-level statistics, filter differential-gene sets and calculate overlapping, unique and shared patterns. Support Paroxetine 2017 notebook-compatible queries when applicable.
scrna-deg-summary
Summarize multi-sheet single-cell RNA-seq differential-gene workbooks, identify cell-type sheets, gene/statistic columns and adjusted-p-value thresholds, count significant genes and query gene significance across immune cell types.
spline-response-modeling
Identify response-curve structures in biological ratio, frequency, dose or proportion tables; fit natural splines and quadratic/cubic regressions, and report peaks, R-squared, F-test p-values, confidence intervals and best-model statistics.
tabular-spreadsheet-io
Inspect CSV, TSV, XLS and XLSX bioinformatics tables, list worksheets, identify headers and wide/long formats, and check sample-identifier alignment before downstream analysis.
No matching items. Try a different keyword.
Tools
alphafold_tool
Query AlphaFold DB for predicted protein structures, confidence scores and model-file information.
archive_dataset_prep_tool
Extract ZIP, TAR and other dataset archives and organize analysis-ready working directories and input files.
cbioportal_tool
Query cancer-study samples, mutations, copy number, expression and clinical data in cBioPortal.
check_sample_overlap
Compare count-matrix column names with metadata sample identifiers to detect missing, duplicate or mismatched samples.
chembl_tool
Query ChEMBL for active compounds, drug targets, assays and bioactivity data.
clinicaltrials_tool
Query ClinicalTrials.gov for trials, conditions, interventions, outcomes, study status and recruitment information.
clinvar_tool
Query ClinVar for variant clinical significance, related conditions, evidence submissions and review status.
crawl_tool
Fetch a specified web page and extract Markdown content for searching and citation.
db_contrast_cache_tool
Cache and retrieve intermediate database-comparison results to reduce repeated remote queries.
dbsnp_tool
Query NCBI dbSNP for rsIDs, alleles, genomic coordinates and variant annotations.
deseq2_results_query_tool
Query DESeq2 result RDS files by gene for log2FoldChange, p-values and adjusted p-values.
encode_tool
Query ENCODE functional-genomics experiments, samples, assays, files and processing metadata.
ensembl_tool
Query Ensembl for genes, transcripts, exons, genomic coordinates and identifier mappings.
execute_in_sandbox
Run shell commands in an isolated environment for analysis, file conversion and artifact generation.
file_info
Read basic file information such as path, type and size without loading the full contents.
gnomad_tool
Query gnomAD for population allele frequencies, filter status and gene-constraint information.
gtex_tool
Query GTEx for cross-tissue gene expression and tissue-specific regulatory data such as eQTLs.
gwas_catalog_tool
Query GWAS Catalog traits, associated loci, risk alleles, studies and statistical results.
infer_table_schema
Infer CSV/TSV delimiters, headers, column counts, data types and wide/long table structure.
jaspar_tool
Query JASPAR for transcription-factor binding motifs, position frequency matrices and species annotations.
kegg_offline_enrichment_tool
Run reproducible pathway overrepresentation analysis using a fixed local KEGG snapshot.
kegg_tool
Query gene, pathway, disease, reaction and compound records through KEGG REST.
list_dir
List directories and files to inspect dataset structure and candidate inputs.
load_memory_tool
Read saved project or run memory by its identifier.
load_recall_candidate_tool
Load the full details of candidates already returned by Memory Recall.
ncbi_blast_tool
Submit sequence-similarity searches to NCBI BLAST and retrieve sequence matches and alignment statistics.
ncbi_search_tool
Query NCBI Entrez for public PubMed, Gene, SRA, GEO and sequence records.
opentargets_tool
Query Open Targets for target–disease associations, genetic evidence and drug-development information.
pdb_fetch_tool
Retrieve structure details or download structure files by PDB identifier.
pdb_search_tool
Search RCSB PDB structures by keyword, sequence or structural criteria.
pubchem_tool
Query PubChem compounds, structures, properties, bioactivity and identifiers.
quickgo_tool
Query QuickGO for GO terms, hierarchies and gene/protein functional annotations.
r_read_rds_tool
Use controlled R scripts to read or inspect RDS object structures and content summaries.
reactome_tool
Query Reactome's curated pathways, reactions, participating molecules and enrichment results.
read_data
Detect common text or table formats and read data previews for subsequent analysis.
read_skill
Read a Skill's description, permitted capabilities, tools and script entry points by name.
read_tool_result
Retrieve or search a previously saved complete tool result by its result identifier.
recall_candidates_tool
Search the current Recall source for previous runs, conclusions or relevant memory candidates.
search_memory_snippets_tool
Search saved memory text for keywords and return concise matching snippets.
search_memory_tool
Use BM25 to search memory titles, summaries and tags for relevant content.
skill_call
Execute capabilities, tools or script entry points after Skill allowlist and runtime permission checks.
stringdb_tool
Query STRING protein interaction networks, functional associations and network enrichment information.
summarize_and_remember_tool
Use a model to summarize long text and persist the summary as searchable memory.
tavily_search
Search the web through Tavily and return relevant pages, summaries and source links.
uniprot_tool
Query UniProt for protein sequences, functions, subcellular locations, synonyms and annotations.
xlsx_reader_tool
List Excel sheets and inspect a selected sheet's headers, structure and data preview.
No matching items. Try a different keyword.
Databases
alphafold
IntegratedAlphaFold DB: predicted protein structures, per-residue confidence and model files.
cbioportal
IntegratedcBioPortal: mutations, copy number, expression, samples and clinical data from cancer cohorts.
chembl
IntegratedChEMBL: drug-like molecules, targets, assays and compound–target bioactivity.
clinicaltrials
IntegratedClinicalTrials.gov: trial protocols, conditions, interventions, outcomes, study status and recruitment information.
clinvar
IntegratedClinVar: clinical significance of human genetic variants, associated phenotypes and submitted evidence.
dailymed
Not yet integratedDailyMed: current FDA drug labeling, ingredients, indications, dosing and safety information.
dbsnp
IntegrateddbSNP: short genetic variants, rsIDs, alleles and genomic locations.
emdb
Not yet integratedEMDB: three-dimensional electron microscopy density maps, including cryo-EM, and experimental metadata.
encode
IntegratedENCODE: transcriptional regulation, chromatin, transcription-factor and functional-genomics experiment data.
ensembl
IntegratedEnsembl: genomes, genes, transcripts, exons, homology relationships and variant annotations.
geo
Not yet integratedGEO: gene-expression, sequencing and other functional-genomics studies and sample data.
gnomad
IntegratedgnomAD: variant frequencies and gene-constraint metrics from large population sequencing datasets.
gtex
IntegratedGTEx: gene expression and expression quantitative trait loci across normal human tissues.
gtopdb
Not yet integratedGuide to PHARMACOLOGY: curated drug targets, ligands and their pharmacological relationships.
gwas_catalog
IntegratedGWAS Catalog: curated trait–locus associations, study samples and significance statistics.
interpro
Not yet integratedInterPro: integrated protein families, domains, functional sites and classification annotations.
iucn
Not yet integratedIUCN Red List: species conservation status, population trends, threats and distributions.
jaspar
IntegratedJASPAR: transcription-factor DNA-binding motifs and position frequency matrices.
kegg
IntegratedKEGG: integrated knowledge of genes, pathways, diseases, drugs, reactions and metabolites.
monarch
Not yet integratedMonarch Initiative: integrated genes, diseases, phenotypes, model organisms and their associations.
mpd
Not yet integratedMouse Phenome Database: mouse-strain phenotypes, genotypes and experimental measurements.
ncbi_blast
IntegratedNCBI BLAST: remote sequence-similarity searches returning sequence matches, alignments and significance statistics.
openfda
Not yet integratedopenFDA: public FDA regulatory and adverse-event data for drugs, devices, food and related areas.
opentarget
IntegratedOpen Targets: integrated target–disease associations, genetics, omics and drug evidence.
pbdb
Not yet integratedPaleobiology Database: fossil taxonomy, geological ages, localities and paleobiological occurrence records.
pdb
IntegratedProtein Data Bank: experimental three-dimensional structures of proteins, nucleic acids and complexes.
pride
IntegratedPRIDE: mass-spectrometry proteomics projects, raw files, identification and quantification results.
pubchem
IntegratedPubChem: compound structures, physicochemical properties, bioassays and linked identifiers.
quickgo
IntegratedQuickGO: Gene Ontology terms, relationships and gene/protein functional annotations.
reactome
IntegratedReactome: curated biological pathways, reactions and participating molecules.
remap
Not yet integratedReMap: transcriptional regulator binding regions integrated from ChIP-seq and related experiments.
stringdb
IntegratedSTRING: known and predicted protein functional associations and interaction networks.
ucsc
Not yet integratedUCSC Genome Browser: genome sequences, annotation tracks and cross-species comparative data.
unichem
Not yet integratedUniChem: compound identifier mappings across chemical databases.
uniprot
IntegratedUniProt: protein sequences, functions, domains, locations and cross-database annotations.
worms
Not yet integratedWoRMS: marine organism taxonomy, scientific names, synonyms and distributions.
No matching items. Try a different keyword.
From results to insights, with AI
Analysis is not the end. Understanding comes next.
High-quality reports
Charts, statistics and methods brought together in one report.
Biological interpretation with AI
Use analysis results and biological knowledge to understand your data.
Trace the complete workflow
Review each analysis step, intermediate output and result.
Export and share
PDF, Word and HTML formats support research discussion and sharing.
Product illustrations and the example report retain their original Chinese text.
See Bioinfoysis at work
Explore research questions, analysis workflows and report delivery.
The video could not be played. Please try again.
Building a prognostic risk model for diffuse large B-cell lymphoma
Case background
Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma and is clinically heterogeneous: even with the same standard chemotherapy regimen (R-CHOP), patient outcomes vary considerably. Prognostic risk models based on gene expression profiles support precise risk stratification and individualized treatment. This analysis uses a published dataset containing microarray expression profiles and complete clinical follow-up information from 420 patients with DLBCL, including molecular subtype (GCB / ABC), survival status and time, stage, and LDH.
Result interpretation
Using transcriptomic data from 420 patients with DLBCL, the analysis built a 63-gene LASSO-Cox risk model that separated patients into high- and low-risk groups (median OS: 1.51 years vs not reached, P=1.18e-31). The risk score remained independent of age, stage, LDH, ECOG, and other traditional clinical factors (adjusted HR=4.59), improving prognostic discrimination from a C-index of 0.731 to 0.862. The model includes known genes associated with the DLBCL immune microenvironment (SPP1 and MS4A4A), suggesting potential value for clinical stratification. Validation in an independent external cohort is still required before use in clinical decision-making.
View full case report Opens in a new windowThe video could not be played. Please try again.
Single-cell pseudotime analysis of hematopoietic differentiation
Case background
Mouse bone marrow single-cell RNA-seq data are used to reconstruct hematopoietic differentiation, identify branch points and explore genes that change along developmental trajectories.
Result interpretation
The supplied report uses PAGA and Diffusion Pseudotime on 8,399 cells to reconstruct six differentiation paths and four branch points, with branch-associated regulators and dynamic gene expression.
View full case report Opens in a new windowThe video could not be played. Please try again.
Protein–protein interaction (PPI) network analysis
Case background
This ulcerative colitis example combines differential expression, protein–protein interaction networks and topology analysis to identify candidate hub genes.
Result interpretation
The supplied report identifies 299 differentially expressed genes and builds a STRING network with 224 nodes and 1,283 edges. MCODE and multiple topology algorithms then identify eight candidate hub genes.
View full case report Opens in a new windowThe video could not be played. Please try again.
Single-cell RNA-seq clustering and annotation
Case background
The PBMC3k example starts with a peripheral blood mononuclear cell expression matrix and covers quality control, normalization, dimensionality reduction, clustering and cell-type annotation.
Result interpretation
The supplied report retains 2,643 cells after quality control, applies Louvain clustering and UMAP visualization, and uses marker analysis to annotate ten clusters as nine PBMC cell types.
View full case report Opens in a new windowThe video could not be played. Please try again.
Biomarker screening with machine learning
Case background
Heart tissue expression profiles from 106 patients with hypertrophic cardiomyopathy and 39 healthy controls are combined with oxidative stress gene sets and machine learning to identify candidate biomarkers.
Result interpretation
The supplied report narrows 899 differentially expressed genes to 23 oxidative stress candidates. Intersecting LASSO and SVM-RFE results identifies PRKCD, MGST1 and CA3, followed by expression comparisons and ROC analysis.
View full case report Opens in a new window