Hymenopteran Venom Dataset Curation

This repository documents the curation process applied to a dataset of venom proteins from Hymenopteran species. The goal was to ensure data quality and prepare the dataset for downstream analyses, such as evolutionary studies and functional annotation.

Curation Overview

The dataset curation process involved the following steps:

Closed Curation Tasks

1. Check for Duplicates:

   • Identified and removed redundant sequences to ensure protein uniqueness.

2. Add Missing Mature Sequences:

   • Filled in gaps by predicting and adding missing mature sequences where needed.

3. Proteins Containing Gaps:

   • Addressed proteins with alignment gaps to improve sequence quality.

4. Check Mature Sequences for Signal Peptides:

   • Used SignalP 6.0 to identify and verify signal peptides.
   • Updated mature sequences based on SignalP predictions.

5. Check Mature Sequences for Propeptides:

   • Used DeepPeptide to identify and verify propeptides.
   • Updated mature sequences based on DeepPeptide predictions.

6. Search publications for toxins not yet in the dataset.

Tools and Software

• SignalP 6.0: Used for signal peptide prediction. https://github.com/fteufel/signalp-6.0
• DeepPeptide: Used for Propeptide prediction https://github.com/fteufel/DeepPeptide
• Others (see scripts).

Repository Structure

├── data/
│   ├── # Dataset in different versions
├── json-files/
│   ├── # Prediction files generated from SignalP6.0 or DeepPeptide
├── logs/
│   ├── # Logs generated from scripts
├── src/
│   ├── # Scripts
├── curated_Dataset.xlsx    # The curated dataset file
├── README.md               # Project documentation (this file)