novoalab / modPhred

modPhred is a pipeline for detection of DNA/RNA modifications from raw ONT data

modPhred

modPhred is a pipeline for detection, annotation and visualisation of DNA/RNA modifications from raw ONT data.

The pipeline consists of four steps / modules:

1. modEncode: Encoding modification probabilities in FastQ (mod_encode.py)
2. modAlign: Build alignments keeping modification information in BAMs (mod_align.py)
3. modReport Extraction of RNA modification information (bedGraph) and QC reports (mod_report.py)
4. modAnalysis: Plotting venn diagrams (mod_plot.py), co-occurrence of modifications (mod_correlation.py) and per-read clustering based on modification profiles (mod_cluster.py)

For more information, please visit full documentation at https://modphred.readthedocs.io.

Does it work with any basecaller?

No, ModPhred works with Guppy flip-flip basecalling models (which can be custom trained, e.g. modification-aware basecalling models trained for example, with taiyaki). Unfortunately, ModPhred does not work with Dorado.

As of March 2024, ONT has deprecated Guppy. Consequently, as of March 2024, ModPhred has also become deprecated. We will update this README if in the near future we upgrade ModPhred to support Dorado or alternative basecallers.

Citation

If you find this work useful, please cite:

Pryszcz LP and Novoa EM (2022) ModPhred: an integrative toolkit for the analysis and storage of nanopore sequencing DNA and RNA modification data. Bioinformatics, 38:257-260.