Msstitch

Added
- FDR calculator from percolator output for PSM tables from MSGF, `msspsmtable percolator`
- Missed cleavage count column in PSM table as an option to `msspsmtable specdata --miscleav`

Changed
- Linear modeled q-values from scores and bestpeptides for a feature are NA if there are fewer than 10 q-values above the threshold.
- Add biological set to PSM table is now optional with `msspsmtable specdata --addbioset` flag.
- Protein FDR calculation does not crash on ZeroDivisionError when T=0: D/T is then set to 1.

Fixed
- Bugs in `msslookup makedecoy` fixed, actually filter out all non-matching peptides and too short peptides
- Protein FDR calculation was for some proteins calculated as D/T+D where that should be D/T
- Peptide and PSM q-value calculation maxes at 1 (in case decoys > targets)

Added
- Trypsinization tool `msslookup trypsinize` to create trypsinized FASTA from proteins
- Biomart table header addition of "Gene stable ID" to comply with latest
- Params `--maxshuffle`, `--minlen`, and `--notrypsin` added to `msslookup seqspace` for more flexible/quicker operation if needed

Changed
- Faster SQL for known_searchspace by UNIQUE constraint, accept text files also instead of .fa,

Added
- Decoy creation tool to do protein-reverse and tryptic peptide reverse while shuffling decoy peptides that are identical to target peptides

Fixed
- For some reason had missed that --noncentric also does not like having a coverage table (dependent on proteins, so irrelevant).

Changed
- Quicker XML parsing with big documents since there was non-cleared elements earlier

Changed
- Not querying SQLite for each feature/MS2 quant in msslookup quant, instead keep all data for given spectra file in memory, hopefully this speeds things up
Fixed
- --noncentric flag for building combined output tables did not work if there was not a protein table in the lookup file. Now possible.

Changed
- Can now build lookups of multiple protein/gene/peptide tables without the nr-of-psms used for isobaric quant, to allow for other quant methods
- Output merged protein/gene tables with all mappings. Whereas protein tables earlier also got gene IDs and gene names, the reverse now is also true
Fixed
- Header names in merged tables are now not only "Protein accession" anymore, but in case of e.g. ENSEMBL gene IDs the table is called gene ID
- Bug fixed, in merged tables we could get eg last line containing a protein with only NAs because it was not checked if the last merged feature was "empty"