MTS-Finder

MTS-Finder is a desktop application for discovering mitochondrial targeting sequences (MTS / presequences) in peptide-level shotgun (bottom-up) proteomics data. It cross-references identified peptides against curated Uniprot presequence annotations and TargetP-2.0 cleavage-site predictions, flags peptides that fall within a mitochondrial presequence, fetches gene symbols from Uniprot, and computes quantitative comparisons between experimental conditions.

Current version: v3.7 (2026) — see CHANGELOG.md for the full release history. v3.6 was the first fully working release; the original v3.4 (2023) was non-functional.

---

Features

• MTS / presequence detection — matches each peptide's master-protein accession against two reference sets:
  • Uniprot curated presequence (transit-peptide) annotations (files/Uniprot_MTS.xlsx)
  • TargetP-2.0 mitochondrial cleavage-site predictions for the MitoCarta3 proteome (files/TargetP2_0_prediction_mitocarta3.xlsx)
• Cleavage-site aware filtering — a peptide is reported only when it lies within the presequence, i.e. its last residue ends at or before the predicted/annotated cleavage site.
• Automatic gene-symbol lookup — resolves accessions to gene symbols via the EBI Proteins API, with a retry and a graceful fallback to the accession ID when offline.
• Flexible input — reads Proteome Discoverer–style peptide exports as either Excel (.xlsx) or tab-delimited text (.txt).
• Normalized or raw abundances — pick normalized or raw channels with a single toggle. Three header styles are recognised, and the two groups are never mixed up even when both are present in the file. A normalized run labels its condition columns WT (Normalized), so the output states which abundances it was built from.
• Channel-to-condition mapping — a simple comma-separated list assigns each abundance channel to an experimental condition; skip and Boost channels are dropped automatically.
• Built-in quantification — for each requested comparison pair it computes:
  • Log2(ratio) of the group means
  • pvalue from an unpaired (Welch-style, equal_var=True) two-sample t-test
  • -Log10 pvalue (volcano-plot ready)
  • Degenerate rows (zero/negative ratios, empty groups) yield NaN instead of crashing the run.
• Excel output — results are written to an .xlsx workbook with one row per matched peptide, and an Open button launches the file when done.
• Simple GUI — a Tkinter interface with file browsers and editable Conditions/Pairs boxes; long-running analysis executes on a background thread so the window stays responsive.

---

Requirements

• Python 3.8+ (developed and verified on 3.12)
• Python packages — all listed in requirements.txt:
  • pandas
  • scipy
  • requests
  • openpyxl (Excel read/write)
• tkinter — powers the GUI, but it ships with the Python standard library rather than PyPI, so it is not in requirements.txt. It is included with the python.org installers on Windows and macOS; on Debian/Ubuntu run sudo apt install python3-tk.
• Internet access (optional but recommended) for gene-symbol lookups via the EBI Proteins API.

---

Installation

git clone https://github.com/science64/MTS-Finder.git
cd MTS-Finder
pip install -r requirements.txt

Using a virtual environment keeps these packages isolated from your system Python:

python -m venv .venv

# Windows (PowerShell)
.venv\Scripts\Activate.ps1

# macOS / Linux
source .venv/bin/activate

pip install -r requirements.txt

Verify the install:

python -c "import pandas, scipy, requests, openpyxl, tkinter; print('All dependencies OK')"

Make sure the files/ folder (shipped with the repository) stays next to main.py — it contains the reference databases and the application icon that the tool loads at runtime.

---

Usage

Launch the application from the project root:

python main.py

Then, in the window:

1. Browse and select your peptide file (.xlsx or tab-delimited .txt).
2. Choose Normalized values or Non-Normalized values.
3. Enter an Output Name for the result workbook.
4. Review/edit the Conditions and Pairs boxes (these are pre-filled from condtions.txt and pairs.txt; you can also load them from a file with their Browse buttons).
5. Click RUN. Progress is shown in the status box.
6. When finished, click Open to view the generated <Output Name>.xlsx, saved next to your input file.

Note: the chosen Conditions and Pairs are also written back as condtions.txt and pairs.txt in the output folder, so your last configuration is preserved.

Conditions

A single comma-separated line with one label per abundance channel, in the same order as the channels appear in the file. Special labels are dropped before analysis:

• skip — ignore this channel
• Boost — ignore the carrier/boost channel

Example:

Light,WT,WT,WT,WT,KO,KO,KO,KO,skip,Boost

Pairs

Semicolon-separated comparisons, each written as numerator/denominator. Each pair produces a Log2, pvalue, and -Log10 pvalue column. For a KO/WT comparison, KO is the numerator (group 2) and WT is the denominator (group 1):

KO/WT

Multiple comparisons:

KO/WT; Rotenone/DMSO; Antimycin/DMSO

Expected input columns

The peptide file should contain the typical Proteome Discoverer peptide-export columns, including:

• Positions in Master Proteins (e.g. P49189 [275-293], or multiple ; -separated accessions)
• Modifications
• Abundance channels, in any one of these three header styles:

Style	Non-Normalized	Normalized
Proteome Discoverer	Abundance: F1: 126, Sample	Abundances (Normalized): F1: 126, Sample
Spaced	Abundance F1 126 Sample	Abundances Normalized F1 126 Sample
Underscore	Abundance_126	Abundances_Normalized_126

A file may contain both groups at once; the toggle selects one group and never mixes the two. Derived Proteome Discoverer columns (Abundance Ratio, Abundances Count, Grouped, Scaled, CV) are ignored — they are not channels. If no channel matching your toggle exists, or the number of channels does not match the number of conditions, the run stops with a message naming the columns it found instead of failing obscurely.

Output columns

Column	Description
Accession	Uniprot accession of the master protein
Gene Symbol	Gene symbol resolved from Uniprot (falls back to accession)
Positions in Master Proteins	Gene symbol with peptide position range
Modifications	Peptide modifications
Uniprot / Uniprot Location	Whether matched via Uniprot annotation, and the cleavage-site position
TargetP / TargetP Location	Whether matched via TargetP-2.0 prediction, and the cleavage-site position
(condition channels)	Abundance values, renamed to your condition labels. On a Normalized run the labels carry a (Normalized) suffix — e.g. WT (Normalized) — so the workbook records which abundances were used. Ratio column names are unaffected.
Log2(num/den)	Log2 ratio of group means, per pair
pvalue(num/den)	Unpaired two-sample t-test p-value, per pair
-Log10 pvalue(num/den)	−log10 of the p-value, per pair

---

Reference data

The files/ directory contains the bundled reference databases used by the matcher:

• Uniprot_MTS.xlsx — Uniprot presequence (transit-peptide) locations
• TargetP2_0_prediction_mitocarta3.xlsx — TargetP-2.0 mitochondrial cleavage-site predictions
• Total_accession_precurser.xlsx — combined accession list used for fast pre-filtering
• icon.ico — application icon

---

Project structure

MTS-Finder/
├── main.py          # Tkinter GUI and analysis orchestration
├── functions.py     # MTS-finding engine, Uniprot/TargetP matching, statistics
├── condtions.txt    # Default conditions line
├── pairs.txt        # Default comparison pairs
├── requirements.txt # Python dependencies
├── files/           # Reference databases and icon
├── CHANGELOG.md
├── LICENSE
└── README.md

---

License

Released under the MIT License. Copyright © 2023 Süleyman Bozkurt. See LICENSE for the full text.

---

Citation / contact

Developed by Süleyman Bozkurt. For questions or issues, please open an issue on the project repository.