How amiraynede/lb2-project-group-1 works
This repository is a reproducible bioinformatics workflow for collecting eukaryotic protein sequences and evaluating signal-peptide classifiers. Its principal path runs from UniProtKB retrieval through redundancy reduction and dataset splitting, then analysis and model evaluation. The graph retains both documented classifier branches and their performance outputs. Several stages are represented by their existing directories because their implementation was not included in the source excerpts; cross-stage links follow the README’s documented workflow rather than inspected calls.
Main entry points: training_sets Training and benchmark data · 00_data_collect.py Sequence retrieval · fasta_cleaner.py FASTA cleaning · output Positive and negative sets
Components 19
Data collection
- Sequence retrieval
01_data_collection/scripts/00_data_collect.py - Positive and negative sets
01_data_collection/output
Dataset preparation
- Sequence clustering
02_data_preparation/files/cluster_output - Representative IDs
02_data_preparation/scripts/01_extract_rep_ids.py - Metadata organization
02_data_preparation/scripts/02_organizing_metadata.py - Cross-validation sets
02_data_preparation/scripts/03_dss_script.py - Training and benchmark data
02_data_preparation/files/training_sets
Data analysis
- Dataset analysis
03_data_analysis/scripts/04_Data_Analysis.ipynb - Dataset visualizations
03_data_analysis/visualization
Classification
- von Heijne training
04_vonHeijne/scripts/06_vonHeijne.ipynb - von Heijne training data
04_vonHeijne/scripts/05_training_df.ipynb - SVM feature extraction
05_SVMs/scripts/07_feature_extraction.ipynb - SVM training
05_SVMs/scripts/08_SVM_main.ipynb
Performance evaluation
- von Heijne evaluation
06_performance/Von_Heijne - SVM evaluation
06_performance/SVMs - Benchmark datasets
06_performance/files - FASTA cleaning
06_performance/scripts/fasta_cleaner.py
Other components
- Researcher
- UniProtKB
Connections 20
- Researcher → Sequence retrieval: initiates
README.md - Sequence retrieval → UniProtKB: queries
01_data_collection/scripts/00_data_collect.py - Sequence retrieval → Positive and negative sets: writes datasets
01_data_collection/scripts/00_data_collect.py - Positive and negative sets → Sequence clustering: cluster sequences
README.md - Sequence clustering → Representative IDs: extract representatives
README.md - Representative IDs → Metadata organization: filter metadata
README.md - Metadata organization → Cross-validation sets: organize datasets
README.md - Cross-validation sets → Training and benchmark data: produce splits
README.md - Training and benchmark data → Dataset analysis: analyze sequences
README.md - Dataset analysis → Dataset visualizations: generate plots
README.md - Training and benchmark data → von Heijne training data: organize training data
README.md - von Heijne training data → von Heijne training: train PSWM
README.md - Training and benchmark data → SVM feature extraction: extract features
README.md - SVM feature extraction → SVM training: train SVM
README.md - von Heijne training → von Heijne evaluation: evaluate classifier
README.md - SVM training → SVM evaluation: evaluate classifier
README.md - Training and benchmark data → Benchmark datasets: benchmark on
README.md - Benchmark datasets → FASTA cleaning: clean FASTA
- FASTA cleaning → von Heijne evaluation: prepare sequences
- FASTA cleaning → SVM evaluation: prepare sequences