The problem. Identifying peptides means matching each observed spectrum against theoretical spectra from a protein database. If you also want to catch post-translational modifications, you must widen the mass tolerance enormously — an “open search” — which explodes the candidate space and made comprehensive searches painfully slow.

The idea. MSFragger uses a fragment-ion indexing scheme that makes spectrum-to-peptide matching dramatically faster, so open searches over huge modification spaces become tractable. Suddenly you can survey all mass shifts in a dataset — discovering unexpected modifications — instead of only looking for the ones you pre-specified.

Why it matters. This is the identification engine in the proteomics stack whose other pieces I’ve already read: MaxQuant/Andromeda and Percolator do related jobs, and MSFragger is the speed-and-openness leap. Conceptually it rhymes with the FM-index/BWA thread from day 6 — the right index turns an intractable search into a fast one. Same algorithmic lesson, proteomics instead of genomics.

Verdict. Foundational for modern proteomics, especially open and modification-focused searches; it anchors the FragPipe ecosystem. Read it for the fragment-indexing trick and for how much biology (unanticipated modifications) becomes visible once the search is cheap enough to run wide.