The following preprocessing steps are performed for candidate compounds with a target (CLK1):

1. Make Fingerprints: This step involves generating 14 fingerprints for the candidate compounds using the provided compound file (data/CLK1.csv), seed file (data/bindingDB_CLK1.csv), compound cutoff (0.7), output path (result/CLK1/), and performing fingerprint generation in parallel.

Note: On a 16 core cpu workstation, fingerprint processing for ~40,000 compounds (cutoff ≥ 0.85) typically takes around 1.9 hours.

time python3 source/0_1_makeFP.py -t data/CLK1.csv -a data/bindingDB_CLK1.csv -c 0.7 -o result/CLK1/ -d

The following steps are performed for network construction:

2. Make Network: In this step, a network is constructed using the provided parameters:

   • Input path: ${sPath}

   • Compound cutoff: ${dCompoundCutoff}

   • Edge cutoff: ${dEdgeCutoff}

   • Seed cutoff: ${dSeedCutoff}

time python3 source/0_2_makeNetwork.py -i ${sPath} -c ${dCompoundCutoff} -e ${dEdgeCutoff} -s ${dSeedCutoff}

The following steps are performed for network propagation:

3. Network propagation: is performed using the seed and edge information.

   • The seed file : {sPath}/{dCompoundCutoff}/seed_{dSeedCutoff}.txt

   • and the corresponding edge files are used for propagation.

sNPSeed=${sPath}${dCompoundCutoff}/seed_${dSeedCutoff}.txt
for sNPEdge in  $(find ${sPath}${dCompoundCutoff} -name edge_${dSeedCutoff}_${dEdgeCutoff}_*); do
    sNPresult=$(echo  $sNPEdge  |  sed  -e  "s/edge/NP/g")
    time python3 source/0_3_network_analysis.py ${sNPEdge} ${sNPSeed} -o ${sNPresult} -addBidirectionEdge True
done

In this final step, network analysis is performed on the propagated network.

The network propagation results {sPath}/{dCompoundCutoff}/NP_{dSeedCutoff}_{dEdgeCutoff}

and the compound file {sPath}/Data_Comp_{dCompoundCutoff}.csv

are utilized.

time python3 source/0_4_HitDiscovery.py -n ${sPath}${dCompoundCutoff}/NP_${dSeedCutoff}_${dEdgeCutoff}_ -f ${sPath}Data_Comp_${dCompoundCutoff}.csv`

./exe.sh 0.85