Chemical probing
Chemical probing experiments, such as SHAPE (Selective 2’-Hydroxyl Acylation Analyzed by Primer Extension) and DMS (Dimethyl Sulfate), measure how reactive each nucleotide of an RNA is to a chemical probe, which shows which parts of the RNA are paired and which are flexible. RNAcentral is bringing these data together: the results are planned to appear in RNAcentral in mid 2027 (release 31), in a new chemical probing tab on sequence pages and as reactivity tracks in the genome browser. This will make RNAcentral a central hub for chemical probing data.
A metadata standard for probing experiments
Chemical probing datasets are often deposited with incomplete or inconsistent descriptions of how they were produced, which makes them hard to find, reuse and compare. To address this, the rnacentral-probing-metadata repository sets out a standard for describing probing experiments. Every dataset is described by a metadata file that follows a shared schema, recording:
the organism (and, for viruses, the strain);
the probing method and chemical, the readout principle and the reverse transcriptase;
a term from the Ontology for Biomedical Investigations (OBI) for the assay, such as OBI:0003886 (NAI RNA structure mapping assay), so the type of experiment is recorded in a standard vocabulary shared with other resources;
the experimental context, such as in vivo or in vitro, and the RNA fraction sequenced;
the publication;
the raw data accessions, with each sample’s group, condition and replicate.
More than 140 SHAPE and DMS datasets are already described this way.
Anyone can add their own dataset by opening a pull request with a new metadata file, following the instructions and template in the repository. Each file is checked automatically against the schema and then reviewed by the RNAcentral team. The raw data must be in a public archive such as ENA, SRA or GEO, and each treated sample group needs at least two biological replicates.
Processing the data
Once a dataset is accepted, its raw data are downloaded and processed with nf-core/rnastructurome, a pipeline developed by the RNAcentral team, against reference sequences and annotations built from RNAcentral, which are kept in the rnacentral-probing-references repository. Where possible, the predicted structures are compared with known structures from Rfam to assess how well the probing data support them.