Proteomics Method Optimization

Systematic comparison of LC gradient lengths and data-dependent acquisition (DDA) Top-N settings using well-characterized reference samples, run on a Thermo Scientific Q Exactive (classic) coupled to an IonOpticks Aurora Ultimate 25 × 75 µm, 1.7 µm analytical column.

Each sub-project targets a specific reference standard and QC question. All notebooks share a common helper library and follow the same structure, so extending the analysis to new raw files requires only dropping files into data/ and re-running the notebook.

---

Sub-projects

Directory	Standard	Purpose
BSA/	Bovine serum albumin tryptic digest (Sigma-Aldrich, P02769)	Baseline sensitivity; single-protein sequence coverage as a focused LC-MS health check
HeLa/	HeLa cell tryptic digest	Complex-matrix benchmarking; whole-proteome identification depth and reproducibility
TMT_Yeast/	Pierce™ TMT11plex Yeast Digest Standard (BY4741)	End-to-end TMT quantification quality; labeling efficiency, ratio accuracy, and differential abundance of three built-in knockouts

---

Repository structure

Method Optimization/
├── README.md
├── helpers/                        # Shared R helper library
│   ├── global.r                    # save_and_show_plot/table, get_method_levels
│   ├── load_data.r                 # load_psm_data()
│   ├── summarize.r                 # PSM counts, coverage, CV, peak sampling, …
│   ├── plots.r                     # plot_metric_distribution, plot_psm_counts, …
│   ├── fwhm.r                      # get_peak_fwhm(), plot_fwhm_vs_rt()
│   └── spectral_quality.r          # get_spectral_quality()
├── BSA/
│   ├── README.md
│   ├── QC_BSA.ipynb
│   ├── data/                       # PD .txt exports + Thermo .raw files (git-ignored)
│   ├── results/                    # Generated CSV tables
│   └── graphs/                     # Generated plots (PNG/PDF)
├── HeLa/
│   ├── README.md
│   ├── QC_HeLa.ipynb
│   ├── data/
│   ├── results/
│   └── graphs/
└── TMT_Yeast/
    ├── README.md
    ├── QC_TMT_Yeast.ipynb
    ├── helpers/
    │   └── tmt_quantification.r    # TMT-specific metrics and plots
    ├── data/
    ├── results/
    └── graphs/

---

Shared helper library

All notebooks source the helpers from ../helpers/ (relative to their own directory). The library provides:

• load_psm_data(filepaths) — reads and merges one or more Proteome Discoverer 3.0.1 PSM export .txt files, parses LC gradient, MS method, and replicate number from the spectrum file name, and removes duplicate scans. Handles multi-node exports (e.g. fixed + dynamic TMT search) by including Identifying.Node.No in the deduplication key.
• summarize.r — PSM counts, unique peptide counts, PSM redundancy, charge distribution, RT distribution, sequence coverage (single-protein or proteome-distribution mode), peptide intensity CV, peak sampling, cycle time.
• plots.r — plot_metric_distribution() (violin + box), bar and density plots for all identification and spectral quality metrics.
• fwhm.r — XIC-based chromatographic peak FWHM via rawrr::readChromatogram(), with per-file RDS caching.
• spectral_quality.r — AGC fill fraction, MS2 TIC, MS2 S/N, and fragment count from Thermo .raw files via rawrr::readSpectrum(), with per-file RDS caching.

---

File naming convention

Spectrum file names must follow this pattern:

<prefix>_<LC_gradient>_<MS_method>_<replicate>.raw

Examples: BSA_90min_Top10_1.raw, HeLa_120min_Top15_3.raw, Yeast_90min_Top10_2.raw

Method metadata (LC gradient length, MS method, replicate number) are parsed automatically. No manual configuration is required when new files follow this convention.

---

Requirements

Easiest setup — JupyterLab Docker Environment

The quickest way to get a fully configured environment with all required packages, JupyterLab, and .NET pre-installed is to use the accompanying Docker image:

SamThilmany/JupyterLab-with-R_Docker-Environment

Manual installation — R packages

Package	Source	Purpose
tidyverse	CRAN	Data manipulation and ggplot2 plotting
patchwork	CRAN	Multi-panel figure assembly
scales	CRAN	Axis label formatting
ggrepel	CRAN	Non-overlapping text labels (volcano plots)
conflicted	CRAN	Explicit resolution of namespace conflicts
IRdisplay	CRAN	Inline HTML display in JupyterLab
rawrr	Bioconductor	Reading Thermo .raw files
MSstatsTMT	Bioconductor	TMT protein summarization and group comparison

Install CRAN packages with install.packages(); install Bioconductor packages with BiocManager::install().

Raw file access

rawrr ≥ 1.15.3 requires the .NET runtime on all platforms (macOS, Linux, and Windows — Mono was removed in 1.15.3). Follow the rawrr installation instructions at https://bioconductor.org/packages/rawrr.

---

Adding new data

1. Place Proteome Discoverer PSM export .txt files in the relevant data/ directory.
2. Place the corresponding Thermo .raw files in the same data/ directory (they are excluded from version control by .gitignore).
3. Re-run the notebook — load_psm_data() picks up all .txt files in data/ automatically.