## Introduction SpecOMS is an Open Modification Search method published originally in David et al. 2017. SpecOMS compares experimental spectra generated by a discovery proteomics experiment to a whole set of theoretical spectra deduced from a protein database in a few minutes on a standard workstation. The procedure yields identification results that might comprise peptides bearing unknown modifications. X!Tandem is a highly regarded classical identification engine, very efficient but unable to find peptide spectrum matches (PSMs) if the precursor ion's mass does not corresponds to the theoretical peptide's mass. Combining the results obtained by running both SpecOMS and X!Tandem reduces the proportion of undecided PSMs compared to a classical approach. However this combination approach raises a number of questions because the two engines do compute PSM scores very differently. We address this problem in this report. ## Methods Our i2MassChroQ quantitative proteomics software has been enhanced by integrating a machine learning algorithm called TIDD (tool-independent and data-dependent machine learning for peptide identification). TIDD is a universal post-processing tool which supports confident peptide identifications regardless of the search engine (Li et al. 2022). i2MassChroQ computes a collection of 30 features describing the PSM (signal:intensity proportion matching theoretical peptide fragments, expected isotope ratio, amongst others). Each PSM is tagged either "target" if the determined peptide sequence was found on the targeted protein database or "decoy" if not. Based on a data subset, TIDD uses a classification Support Vector Machine (SVM) algorithm to learn how to use the features to confidently classify each PSM in the "target" or "decoy" category. TIDD computes an SVM probability that i2MassChroQ uses as a new PSM score in order to filter, group and infer peptides and proteins in the sample. ## Results We have elaborated an efficient tool to merge identification results from different identification engines. Depending on the sample, the number of PSMs in an MS run identification may be greatly increased. As i2MassChroQ is able to handle results from various identification engines (pepXML, mzIdentML, Mascot dat formats are supported), the procedures described above might be applied using other identification engines.