Contact: boileau@uni-heidelberg.de (Etienne Boileau, Dieterich Lab, Klaus Tschira Institute for Integrative Computational Cardiology, Heidelberg University Hospital)
Sci-ModoM is an open source collection of RNA modification sites from high-throughput whole-transcriptome sequencing. This hub exposes those sites as browser tracks for human (hg38) and mouse (mm39).
Datasets are organized in a searchable, faceted table that mirrors the Sci-ModoM Search interface, with facets for modification, biosample (cell line, tissue, or organ) and detection technology. Every dataset carries a permanent EUFID (Epitranscriptome Unified Format identifier), which links back to its full record in the Sci-ModoM Browse interface.
The modifications represented are N6-methyladenosine (m6A), 5-methylcytidine (m5C), pseudouridine (Y), and the 2'-O-methylated nucleosides Am, Cm, Gm and Um.
Datasets are selected on the track configuration page. Use the facet panels on the left to narrow the table, then check the datasets you want to display. The Active tab lists only the datasets currently selected. Because a typical dataset covers only a small fraction of the transcriptome, subtracks with no sites in the current window are hidden automatically. Modification sites are sparse, they may be associated with biosamples, conditions, or technology, so many loci will show no sites at all.
Each item is a single modified position (or a multi-base feature e.g. PRAISE), drawn on the strand it was called on. In a few cases e.g. PsiNanopore, the source study did not provide enough information to assign a strand, so unstranded rows are possible. Items are colored on a continuous blue-to-red ramp by frequency (also called percent modified):
| Color | Frequency |
|---|---|
| lowest reported frequency | |
| 50% modified | |
| 100% modified |
The minimum reported frequency differs between datasets.
Hovering over an item shows the modification, the coverage and the frequency (percent modified). Clicking an item opens a details page with the same values and a link to the Sci-ModoM record for that dataset. The original score value may also be present (see Methods). Each dataset's own settings page, accessible by clicking a dataset name once it is turned on, offers Minimum coverage and Frequency (percent modified) filters, which hide sites outside the chosen range; both default to showing everything.
This track hub was generated using scimodhub. All sites were retrieved from the Sci-ModoM API. For the browser, each dataset was converted to a bigBed file with nine standard BED fields plus two extra fields (a third extra field may be present to store the original bedRMod score), using the following schema:
table bedRMod "RNA modification sites" ( string chrom; "Chromosome" uint chromStart; "Modification start position" uint chromEnd; "Modification end position" string name; "Modification (MODOMICS) short name" uint score; "Not used" char[1] strand; "Strand" uint thickStart; "Thick start" uint thickEnd; "Thick end" uint reserved; "Item color, blue (low) to red (100) percent modified" uint coverage; "Coverage" float frequency; "Percent modified" uint rawScore; "Score as reported in the source bedRMod file" )
For an explanation of the bedRMod score, consult the latest euf-specs. See also here for an explanation how the bedRMod score is handled when generating these tracks.
Sci-ModoM imports published datasets in bedRMod (euf-specs) format, harmonizes them to a common assembly and annotation, and assigns each a permanent EUFID. Refer to the metadata of each dataset via its EUFID for dataset-specific source study information and methodology. Consult the online documentation and the Sci-ModoM publication below for general information, import, liftover and quality control procedures.
The underlying bigBed files can be downloaded from the hub directory at
trackhub.dieterichlab.org/eboileau/SciModHub, and inspected or converted to text with the UCSC utilities
bigBedInfo and bigBedToBed, available from
the UCSC downloads server. For example:
bigBedToBed https://trackhub.dieterichlab.org/eboileau/SciModHub/hsapiens/hg38/jaSvBXBkn3xs2000000090A.bb \
-chrom=chr1 -start=0 -end=1000000 stdout
The same data are available in bedRMod format, together with the full metadata for every dataset, from the Sci-ModoM web interface and its API. Hub tracks can also be queried interactively with the Table Browser and the Data Integrator.
Sci-ModoM is developed and maintained by Etienne Boileau from the Dieterich Lab at Heidelberg University Hospital. Our work is done in collaboration with the Human RNome Project. Thanks to Jonathan Casper and Maximilian Haeussler at UCSC for advice on hub structure and on the faceted track selection interface.
Boileau E, Wilhelmi H, Busch A, Cappannini A, Hildebrand A, Bujnicki JM, Dieterich C. Sci-ModoM: a quantitative database of transcriptome-wide high-throughput RNA modification sites. Nucleic Acids Res. 2025 Jan 6;53(D1):D310-D317. PMID: 39498498; PMC: PMC11701610