parylation-predictor 3.1.0

ParylationPredictor – predict PARylation sites and detect PAR-binding domains in proteins

ParylationPredictor

ParylationPredictor predicts PARylation (poly-ADP-ribosylation) sites in proteins and detects PAR-binding domains. You give it a UniProt ID — it gives you publication-quality figures, a formatted Excel file, an interactive HTML report, and structured data exports, all in one command.

Developed by the Hashemi Gheinani Lab, Department of BioMedical Research, University of Bern.

---

The quickest way to try it — no installation needed

Go to gheinani.github.io/parbm-detector/tool.html, type your UniProt ID (for example Q9NTX7), and click Analyze. All 8 analysis panels appear in the browser and you can download a ZIP with the full results. No Python, no installation, no command line.

Use the Python package when you need:

• Batch analysis of many proteins at once
• Full Excel workbook output with formatted sheets
• The complete 8-panel composite figure as PNG/PDF/SVG for a paper

---

Installing and running the Python package

What you need before starting

• A Mac or Windows PC
• Python 3.8 or newer already installed (see below if you are not sure)
• An internet connection

You do not need to know how to program. You only need to type a few commands exactly as written.

---

Step 0 — Check that Python is installed

On Mac: open the Terminal application.

• Press Command + Space, type Terminal, press Enter.

On Windows: open Command Prompt.

• Press the Windows key, type cmd, press Enter.

In the window that opens, type the following and press Enter:

python --version

You should see something like Python 3.11.2. If you see a version number starting with 3, you are ready. If you see an error or a version starting with 2, visit python.org/downloads and install the latest version before continuing.

---

Step 1 — Download the package

You do not need Git. Just download the package as a ZIP file directly from GitHub:

1. Go to github.com/gheinani/parbm-detector
2. Click the green Code button near the top right
3. Click Download ZIP
4. Once downloaded, find the ZIP file (usually in your Downloads folder) and double-click it to unzip it
5. You will now have a folder called parbm-detector-main

---

Step 2 — Open a Terminal in the right folder

You need to tell the Terminal to look inside the folder you just unzipped.

On Mac:

1. Open Terminal (press Command + Space, type Terminal, press Enter)
2. Type cd (that is: cd followed by a space — do not press Enter yet)
3. Open Finder, find the parbm-detector-main folder, and drag the folder onto the Terminal window — the path will be filled in automatically
4. Press Enter

On Windows:

1. Open the parbm-detector-main folder in File Explorer
2. Click in the address bar at the top of the window (it shows the folder path)
3. Type cmd and press Enter — a Command Prompt will open already pointing to that folder

You can confirm you are in the right place by typing dir (Windows) or ls (Mac) and pressing Enter. You should see files like pyproject.toml and README.md listed.

---

Step 3 — Install the package

In the Terminal / Command Prompt window, type the following and press Enter:

pip install .

Wait for it to finish — it will download and install everything automatically (requests, matplotlib, numpy, openpyxl, plotly). This takes about 1–2 minutes depending on your internet connection. When you see the prompt reappear with no error message, the installation is complete.

If you see "pip: command not found", try pip3 install . instead.

---

Step 4 — Run your first analysis

You will write a short script — a plain text file with the instructions for the tool. Do not worry, you only need to change one line.

1. Open Notepad (Windows) or TextEdit (Mac) — any plain text editor
   • On Mac: if TextEdit opens in rich-text mode, go to Format > Make Plain Text
2. Copy and paste the following text exactly:

from parbm_detector_pkg import PARBMDetector

detector = PARBMDetector()
result = detector.analyze("Q9NTX7")        # <-- replace Q9NTX7 with your UniProt ID
result = detector.enrich_result(result)
folder = detector.export_to_folder(result, base_dir=".")
print("Done! Your results are in:", folder)

3. Replace Q9NTX7 with your UniProt accession ID (for example P18887)
4. Save the file as run_analysis.py
   • On Windows: in the Save dialog, change "Save as type" to "All Files" and type the filename as run_analysis.py
   • On Mac: save as run_analysis.py and make sure TextEdit does not add .txt
5. Save the file inside the parbm-detector-main folder

Now go back to the Terminal / Command Prompt and type:

python run_analysis.py

Press Enter. The tool will connect to UniProt, fetch your protein, run the analysis, and print a progress log. When it finishes, it prints:

Done! Your results are in: ./Q9NTX7_RNF146

If you see "python: command not found", try python3 run_analysis.py instead.

---

Step 5 — Find your results

Open the parbm-detector-main folder in Finder (Mac) or File Explorer (Windows). You will see a new folder named after your protein, for example Q9NTX7_RNF146. Open it — it contains:

Q9NTX7_RNF146/
├── Q9NTX7_RNF146_analysis.png      <- 8-panel figure (open with any image viewer)
├── Q9NTX7_RNF146_analysis.pdf      <- same figure as PDF, ready for a journal
├── Q9NTX7_RNF146_analysis.svg      <- scalable vector version
├── Q9NTX7_RNF146_report.xlsx       <- Excel workbook with 4 data sheets
├── Q9NTX7_RNF146_report.html       <- interactive report (open in any browser)
├── Q9NTX7_RNF146_data.json         <- raw data in JSON format
├── Q9NTX7_RNF146_sites.tsv         <- table of predicted sites (open in Excel)
└── panels/                          <- each of the 8 panels as separate files
    ├── *_A_domain_architecture.png
    ├── *_B_confidence_landscape.png
    ├── *_C_hydrophobicity.png
    ├── *_D_charge_profile.png
    ├── *_E_motif_breakdown.png
    ├── *_F_residue_composition.png
    ├── *_G_disorder_profile.png
    └── *_H_residue_type_breakdown.png

Double-click Q9NTX7_RNF146_report.html to open the interactive report in your browser. Double-click the .xlsx file to open it in Excel. The .png and .pdf files open in any image or PDF viewer.

---

Analyzing multiple proteins

To analyze several proteins at once, replace the analysis lines in run_analysis.py with a list:

from parbm_detector_pkg import PARBMDetector

detector = PARBMDetector()

proteins = ["Q9NTX7", "P18887", "Q9Y6K9"]   # add as many UniProt IDs as you like

for uid in proteins:
    result = detector.analyze(uid)
    result = detector.enrich_result(result)
    detector.export_to_folder(result, base_dir=".")
    print("Finished:", uid)

Each protein gets its own output folder.

---

What does the tool analyse?

Capability	Details
PAR-binding domain detection	Queries InterPro for 11 domain families: WWE, Macro, MacroD-type, BRCT, PBZ, RRM, CCCH, PARP catalytic and regulatory
PARylation site prediction	Scores every residue for SxxE, SxxD, TxxE, ExxE, and acidic-cluster motifs
Disorder prediction	Per-residue IUPred2A scores to find modification-accessible flexible regions
Cross-database enrichment	Known PTMs from PHOSPHO.ELM, pathways from Reactome and KEGG, literature from EuropePMC
Publication figures	8-panel composite PNG/PDF/SVG plus 8 individual panel files
Interactive report	Self-contained HTML report viewable in any browser
Structured exports	Formatted Excel (4 sheets), JSON, and TSV

---

Figure panels explained

Panel	What it shows	How to interpret it
A	Domain architecture	Protein backbone with PAR-binding domain blocks. Sites inside domain boxes are highest-priority candidates
B	Confidence landscape	Sliding-window score across the sequence. Broad peaks indicate modification hotspots
C	Hydrophobicity	Kyte-Doolittle profile. Sites in hydrophilic dips are most solvent-accessible
D	Charge profile	Local net charge. Deep acidic (negative) regions are primary modification zones
E	Motif breakdown	Counts per motif type (SxxE, SxxD, TxxE, ExxE, acidic cluster)
F	Residue composition	Amino-acid pie chart with key metrics summary
G	Disorder profile	IUPred2A scores. Sites in disordered regions (score > 0.5) are most accessible
H	Residue-type breakdown	All predicted vs. high-confidence sites by amino acid (S, T, E, D, Y, K, R)

---

Troubleshooting

"pip is not recognized" or "python is not recognized" Python may not be added to your system PATH. On Windows, reinstall Python from python.org and make sure to tick the box "Add Python to PATH" during installation.

"No module named parbm_detector_pkg" You are running python from the wrong folder. Make sure your Terminal is inside the parbm-detector-main folder (Step 2) before running the install and analysis commands.

The script runs but I get a network error The tool needs internet access to contact UniProt, InterPro, and other databases. Check your connection and try again. Corporate or university firewalls occasionally block API calls — try from a different network if the problem persists.

The analysis takes a long time Each protein analysis makes several API requests. 30–60 seconds per protein is normal. Enrichment (enrich_result) adds another 30–60 seconds because it queries EuropePMC, Reactome, and KEGG.

---

Requirements

• Python 3.8 or newer
• pip (comes with Python)

Python packages installed automatically by pip install .:

Package	Purpose
requests	API calls to UniProt, InterPro, IUPred2A, PHOSPHO.ELM, EuropePMC, Reactome, KEGG
matplotlib	8-panel publication figures
numpy	Sliding-window scoring
openpyxl	Formatted Excel workbook
plotly	Interactive HTML report

---

Contact

Dr. Ali Hashemi Gheinani Group Leader Department of BioMedical Research, University of Bern Email: [e-mail] Lab: Hashemi Gheinani Lab

For bug reports and feature requests, please use the GitHub Issues tracker.

---

License

MIT License — see LICENSE for details.