Large-scale SMILES Requesting and Fingerprints Converting

Written by Jiankun Lyu, 20170918

The hierarchy of the directories:

smiles_requesting/----- working/ 
              |                |
              |                |------ extract_all.sort.uniq.txt file(soft link)
              |                | 
              |                |------ db_zincid/
              |                                                 
              |                                                 
              |
              ------- scripts/ ------ submit.csh
                              |
                              |------ make_chunks_for_file_new.py
                              |
                              |------ setup_converting_fps_files.py
                              |
                              |------ combine_smi_and_fp.py
                              |
                              |------ check_outputs.csh

This tutorial is for requesting a large number of SMILES for docking results from ZINC server. Usually, the number is larger than 5M ZINC IDs.

1) make directories and copy scripts

mkdir smiles_requesting
cd smiles_requesting
mkdir working
mkdir scripts
cd working
mkdir db_zincid
ln -s /path/to/extract_all.sort.uniq.txt
cd ../scripts
cp /mnt/nfs/home/jklyu/zzz.script/large_scale_docking/cluster_analysis/best_first_clustering/converting_fps/submit.csh .
cp /mnt/nfs/home/jklyu/zzz.script/large_scale_docking/cluster_analysis/best_first_clustering/converting_fps/setup_converting_fps_files.py .
cp /mnt/nfs/home/jklyu/zzz.script/large_scale_docking/cluster_analysis/best_first_clustering/converting_fps/combine_smi_and_fp.py .
cp /mnt/nfs/home/jklyu/zzz.script/large_scale_docking/cluster_analysis/best_first_clustering/converting_fps/check_outputs.csh .
cp /mnt/nfs/home/jklyu/zzz.script/large_scale_docking/cluster_analysis/best_first_clustering/converting_fps/make_chunks_for_file_new.py .
cd ../

2) Get ZINC ID and energy columns from the extract_all.sort.uniq.txt file and split the zincid file

cd working/db_zincid
head -(number) ../extract_all.sort.uniq.txt | awk '{print $3" "$22}' > extract_all.top(number).sort.uniq.zincid.energy note: change number in the brackets
python ../../scripts/make_chunks_for_file_new.py extract_all.top(number).sort.uniq.zincid.energy top(number).zincid 500 .
cd ../

3) Create a zincid.sdi file

ls /full/path/to/db_zincid/top(number).zincid_* > zincid.sdi

4) Set up requesting files and directories

python ../scripts/setup_converting_fps_files.py . converting_fps_ zincid.sdi 500 count

5) Submit requesting and converting jobs

csh ../scripts/submit.csh

6) Check if every job finishes

csh ../scripts/check_outputs.csh (number_of_jobs) (prefix)
if you find any missing files, please edit dirlist in the working directory and resubmit them.

7) Collect data from the compressed files

cd db_zincid
python ../../scripts/combine_smi_and_fp.py