Homologous genes & gene families
Homologous gene identification, gene family analysis, and synteny analysis are staples of plant comparative genomics. The platform packages them as one-click pipeline tools in the tool catalog: fill in species sequences and annotations once, and the platform chains the multi-step task chain automatically with intermediate files flowing through (see Manual mode).
Homologous gene identification: 3 approaches
| Approach | Tool | How it works | Best for |
|---|---|---|---|
| One-click pipeline | homolog_one_step | Automatically chains sequence search + RBH/cRBH + OrthoFinder orthogroups + MCScanX synteny | Getting the full set of homology relations plus synteny evidence in one go |
| RBH reciprocal best hits | homolog_rbh | Two species are each other’s best hits (BLAST/DIAMOND); cRBH adds a second-best constraint | Two species, fast and direct ortholog pairs |
| Orthogroup clustering | homolog_orthogroup | OrthoFinder multi-species clustering infers orthologous/paralogous groups | Comparative analysis across 3 or more species |
Inputs and outputs
| Approach | Inputs | Outputs |
|---|---|---|
| RBH | Protein FASTA for two species; e-value / coverage thresholds; optional cRBH alpha | orthologs.tsv (homolog pairs), rbh_detail.tsv (bidirectional hit details), rbh_scatter.png (e-value scatter plot), summary.json |
| orthogroup | Protein FASTA for ≥2 species (file names serve as species names); optional species tree in Newick (-s) | orthogroups.tsv (long form), orthogroups_wide.tsv (species × group matrix), single-copy gene set, OrthoFinder result directory |
| One-click pipeline | Per-species protein/CDS FASTA + genome FASTA + GFF annotation (needed for the synteny step) | Everything from RBH + orthogroup, plus interspecies syntenic blocks |
[Screenshot coming: homolog_one_step form and result file list]
Gene family analysis: a multi-stage task chain
Gene family analysis (identifying members of families like WRKY or NBS-LRR, then running downstream evolutionary analyses) executes as a task chain — one submission runs every stage automatically:
Input preparation → family definition parsing → candidate search (HMM / similarity search)
→ domain validation → isoform resolution → member confirmation
→ protein property analysis → sequence alignment → motif analysis → phylogenetic tree → summary report
- Inputs: protein FASTA for the target species (required); family definition (HMM profile, domain composition, or a list of known members, depending on the family type);
- Domain validation stage: the candidate union is screened against the full Pfam library with hmmscan —cut_ga; isoform representatives are chosen by a deterministic rule chain, with no second E-value threshold;
- Outputs: the family member list (with domain validation results), protein property table (molecular weight / isoelectric point, etc.), multiple sequence alignment, motif patterns, tree figures, and a summary report — archived in per-stage subdirectories;
- Single-stage reruns: every stage is also an independent stage tool; point it at the completed stage directory to rerun just one step (say, redrawing the tree) without rerunning the whole chain.
[Screenshot coming: gene family task-chain stage progress with motif and tree results]
Synteny analysis (MCScanX)
Synteny analysis identifies duplicated blocks and genome structural variation between (or within) species:
- Inputs: CDS/protein sequences + GFF annotation for two (or more) species (the one-click pipeline prepares MCScanX’s .blast and .gff inputs automatically);
- Execution: all-against-all protein comparison → anchor pairing by chromosomal position → MCScanX detects syntenic blocks;
- Outputs: the syntenic block file (.collinearity, with paired genes and block classification: orthologous / paralogous) and the anchor pairing table; usable for Ka/Ks calculation (duplication dating) and genome evolution narratives.
In the homologous gene use case, synteny routinely cross-validates RBH / orthogroup results. The DupGen flow in synteny defaults to the B73v4-NCBI RefSeq pairwise resources.
[Screenshot coming: interspecies synteny circle or dot plot]
Practical tips
- Unify annotation versions first: every species in a comparison must use FASTA + GFF from the same release; mixing versions breaks ID matching;
- Prefer protein sequences: RBH and orthogroup work best on protein FASTA (more tolerant of codon-usage differences than CDS); only the synteny step needs CDS and genomic coordinates;
- Many species → orthogroup, two species → RBH: for speed and directness pick RBH/cRBH; for complete ortholog/paralog structure pick OrthoFinder; for the full evidence package use the one-click pipeline;
- Results are re-usable: homolog pairs and single-copy gene sets are standard tables you can download for tree building, candidate-gene evidence integration, or cross-species transfer of GWAS candidate genes.