We have integrated semi-supervised PSM rescoring in the ANN-SoLo open modification spectral library search engine.

Open modification searching = identify modified peptides without explicitly having to specify PTMs.

PSM rescoring = more accurate and sensitive spectrum identification results.

Open modification searching + PSM rescoring = Win! 💪

Now published in JPR: https://pubs.acs.org/doi/10.1021/acs.jproteome.2c00616
Try ANN-SoLo: https://github.com/bittremieux/ANN-SoLo

PSM rescoring was originally introduced by Percolator, and is increasingly being used to post-process spectrum annotation results from standard searching. But does it also work with open modification searching results?

Yes! We identify a large number of additional modified peptides, including from phosphoproteomics and glycoproteomics datasets.

PSM rescoring is transparently integrated into ANN-SoLo using the mokapot package by @wfondrie.

Bonus: ANN-SoLo is a spectral library tool, allowing us to use PSM features derived from fragment intensities directly. No need for complex deep learning-based spectrum predictions as used by alternative tools.