# lfq — MHCquant(1) spike-in LFQ benchmark

Label-free **quantification accuracy** benchmark for immunopeptidomics.

## Source
- **Publication:** Bichmann et al., *MHCquant: Automated and Reproducible Data Analysis for Immunopeptidomics*, J. Proteome Res. (2019). doi:10.1021/acs.jproteome.9b00313
- **PRIDE / ProteomeXchange:** PXD011628
- **Instrument:** Thermo (Orbitrap), DDA `.raw`

## Design
A defined pool of synthetic tumor-associated peptides (TUMAPs) was spiked into a constant JY cell-line
immunopeptidome matrix at four concentrations, in triplicate, to assess quantification accuracy:

| concentration | replicates | files |
|---|---|---|
| 100 amol | 3 | ...TUMAPs_100am...msms43/44/45 |
| 1 fmol   | 3 | ...TUMAPs_1fm...msms46/47/48 |
| 10 fmol  | 3 | ...TUMAPs_10fm...msms49/50/51 |
| 100 fmol | 3 | ...TUMAPs_100fm...msms52/53/54 |

12 unique `.raw` files (~4.9 GB). Original PXD011628 filenames retained.
(The source set additionally contained renamed `JY_<conc>_1_<rep>.raw` aliases that were
md5-identical to these 12 originals — deduplicated here.)

## Spiked isotope-labeled peptides (66)

**66 isotope-labeled synthetic HLA class I peptides**, each carrying **exactly one** heavy
stable-isotope-labeled residue, spiked into the JY matrix at 0.1 / 1 / 10 / 100 fmol.

Heavy labels (from the labeled residue; mass-verified):
- **Valine** → `Label:13C(5)15N(1)` = **UNIMOD:268** (+6.0138 Da) — 36 peptides
- **Proline** → `Label:13C(5)15N(1)` = **UNIMOD:268** (+6.0138 Da) — 16 peptides
- **Leucine** → `Label:13C(6)15N(1)` = **UNIMOD:695** (+7.0172 Da) — 14 peptides

**Source (authoritative):** MHCquant paper supplementary, Bichmann et al. 2019, *J. Proteome Res.*
(doi:10.1021/acs.jproteome.9b00313), **Supplemental Table S3 "Isotope-labeled peptides"** — the
labeled residue is shown in **bold** per sequence; extracted here via PDF font (Times-Bold) detection.

Full annotated list in `spiked_peptides_labels.tsv` (sequence, labeled residue+position,
UniMod, ProForma, source protein, protein position, MHC restriction).

This 66-peptide set is the superset of the 62-peptide panel in the sister immatics/Tübingen paper
(Nelde/Löffler et al., *Mol Cell Proteomics* 2020, PMC7050110, Table S1); the 4 additional peptides
here are **APRPASSL, KTKGPLKQK, STSSHSLQK, VPRAKIHSL**. (The 3 `*_syn_HCD` reference-spectrum
peptides in PXD011628 — ETLKPGTCVKR, FVPPTAISHF, SRFLSQLDK — are a separate identification set, not
part of this quantification panel.)