Sign in
SigRepo signature repository
Welcome to SigRepo
A snapshot of the repository.
Total signatures
Active users
Organisms
Assay types
By organism
Distribution across organisms
By assay
Counts per assay type
Top contributors
Most active by visible signatures
Signatures
Browse the repository, then open a signature's metadata, features and differential expression.Highlight one or more rows to add them to the basket. The most recently clicked row becomes the active selection, and View loads its full contents on demand.
Selected signature
Selecting a row updates the active signature. Use View to load the full metadata and data tables below.
Collections
Browse the repository's collections, then open one to review its metadata and member signatures.Select a collection row to make it active. Use View to load the full collection details below.
Selected collection
The active collection summary appears immediately. Use View to fetch the collection details and member signatures on demand.
Annotate signatures
Enrich signatures against genesets with hypergeometric, KS or GSEA (fgsea) tests usingSigRepo::runHypeR()
, then explore the results and get the R code that reproduces them.
Step 1
An OmicSignature, a named list of them, or an OmicSignatureCollection. List names become the query labels.
Gene symbols missing from a signature are looked up in the repository's reference tables.
Start each list with '# name'. Genes can be separated by commas, spaces or new lines.
Lines of 'GENE score' make a ranked list, used in the order given: the KS test accepts plain or ranked lists,
GSEA needs ranked lists and the hypergeometric test plain ones.
Files: GMT, a CSV/TSV with gene (and optional score and list) columns, or text in the same format.
Signatures
Pick up to 10 signatures. Picks are kept by signature, so they stay when the filters change.
Step 2
The organism of every signature in the run, and the species MSigDB collections are loaded for.
A run tests one organism: signatures of another organism cannot run together.
The collections offered are the ones in this server's MSigDB cache for the organism.
The largest (C2, C5) are left out: a run against them needs more memory than the server has.
For anything else, load custom genesets.
Organism and genesets
Or type a single geneset
Step 3
Auto uses each signature's measured difexp genes for the hypergeometric test when the difexp looks complete,
and 23467 with a warning when it looks filtered. KS and GSEA rank the whole table, so a population size
does not change them; a gene universe or difexp background reduces queries and genesets to those genes.
Every other runHypeR() argument stays at its default: the ranked table is the difexp, ranked by the score
column and tested at both ends, with power 1, fgsea's own sampling and seed, FDR pooled across the run, and
every geneset kept (filter the plots instead).
Test
Background
Step 4
Preview and run
Results
Run an enrichment to see the dot plot, enrichment curves, maps, tables and the equivalent R code.
Every query's hypeR table. Select a row to open its enrichment plot.
Copy, CSV and Excel export the rows that pass the column filters, at full precision.
The result rendered by hypeR::rctbl_build(): it is hypeR's own hyp/multihyp, so every hypeR function works on it.
The same run and the plots as currently set, from R.
Compare signatures
Feature overlap or rank-based enrichment (KS and GSEA) withSigRepo::compareSignatures()
, all against each other or a first list against a second.
Step 1
An OmicSignature, a list of them, or an OmicSignatureCollection.
An OmicSignature, a list of them, or an OmicSignatureCollection.
Signatures
List 1
Each signature here is compared against every other one, or against List 2 when it is switched on. List 1 is the rows of every result.
List 2
The columns of every result. For KS and GSEA this is the ranking side: its signatures must be bi-directional with a difexp table.
Step 2
KS and GSEA rank by the p-value column. A signature without it is ranked by its pvalue column,
or by the adjusted p-value column if it has neither; the comparison warns when that happens.
Comparison
GSEA
Advanced: background, columns, label pairing
Label pairing
Levels are paired by position (level 1 with level 1, level 2 with level 2), using each signature's own group_label order unless you set both levels here. Uni-directional signatures have no levels to pair.
Step 3
Run
Results
Run a comparison to see the heatmap, tables, and the equivalent R call.
Every signature pair with every measure. Self-comparisons list each overlap pair once.
Copy, CSV and Excel export the rows that pass the column filters, at full precision.
Rows are List 1 (S1, S2, ...); columns are List 2 (R1, R2, ...), or List 1 again for a self-comparison.
The key below gives each label's signature. Exports use the full signature names.
Key
The group label each signature contributes at each level.
The same comparison from R. The result download can be passed straight to OmicSignature::signature_similarity_heatmap().
Reference browser
Search the transcriptomic and proteomic reference features, then inspect a record.Search filters
Choose an organism and assay, then optionally narrow the results to one feature name.
Search results
Selected feature
Selecting a row loads the feature's full record below.