matlab

WINNER — MATLAB implementation

The original MATLAB source for Nguyen et al. (Front. Big Data 2022).

How to run

After formatting your input files in the same way as NeonatalHeartCaseStudy/, open MATLAB and run either RunWinner.m or RunWinner_withPValue.m. Both read the same input files.

• RunWinner.m — simpler and faster; emits a WINNER score per gene but no p-values.
• RunWinner_withPValue.m — also emits expansion and ranking p-values but is substantially slower (10 000 random-network null). Should be run under the supervision of a trained bioinformatician — see the header comments in that file for the extra setup required (AllGeneGloDeg.txt, correct UniProt-style IDs, etc.).

Input files

File	Columns
GeneList.txt	Gene, IsSeeded (S = seed, E = expansion candidate)
Interaction.txt	node1, node2, combined_score ∈ [0, 1]
AllGeneGloDeg.txt	gene id, global degree (p-value mode only)

Output files

• winnerResult.txt — three columns: gene name, seed-or-expand flag, WINNER score (higher = more important).
• winnerResult_withPVal.txt — four columns: gene name, WINNER score, expansion p-value (NaN for seed rows), ranking p-value.

Files in this folder

matlab/
├── spinnerIteration.m                      # core personalized-PageRank kernel
├── pVal_combineUpdate.xlsx                 # supplementary tables
└── NeonatalHeartCaseStudy/                 # worked example
    ├── RunWinner.m                         # simple mode entry point
    ├── RunWinner_withPValue.m              # p-value mode entry point
    ├── spinnerIteration.m
    ├── sym_generate_srand.m                # degree-preserving network randomiser
    ├── GeneList.txt
    ├── Interaction.txt
    ├── AllGeneGloDeg.txt
    ├── winnerResult.txt                    # reference simple-mode output
    └── winnerResult_withPVal.txt           # reference p-value-mode output

See the top-level README for the MATLAB-vs-Python decision guide, and the python/ directory for the Python port.