Skip to content

README.md

Latest commit

 

History

History
96 lines (71 loc) · 3.74 KB

README.md

File metadata and controls

96 lines (71 loc) · 3.74 KB

Generate the data needed

generate_peptides.py

  1. makes sdf files of all combinations of n aminoacids (n=3)
python3 generate_peptides.py

Make a small benchmark of 20 and plot the outputs

compare_diffdock.py

  1. compares rank1_confidence scores for pairs of inference_steps and samples_per_complex (Output:test2.csv) This, was a small benchmarking script to explore these two parameters
python3 compare_diffdock.py
  1. Generates a 3D scatter plot for confidence_score vs samples_per_complex vs inference_steps (Input:dd_test2.csv, output:graph)
python3 plot_3d.py
  1. Plots a graph of samples_per_complex vs confidence score. Can be used for inference_steps by changing the y-axis. (Input: dd_test2.csv, Output: graph)
python3 plot_surface.py

DiffDock analysis

  1. runs diffdock for all n peptides (n=8000)(Output: mdm2_results, dd_fin.csv) This is the main script that does that stuff
python3 diffdock_final.py > dd_fin.csv
  1. UNUSED Calculates the minimum distance between the heavy atoms of p53 and the peptide. Specify the peptide as the input. (Input: peptide directory, output: dd_distance.csv)
python3 dd_distance.py > dd_distance.csv
  1. Calculates the mean and standard deviation for the distance between the ranks (Input: peptides from fin_results, output:dd_stats2.csv)
python3 dd_rank.py > dd_stats2.csv
  1. calculates the maximum length of the peptides (Input: rank1_confidence of peptides, Output: pep_length.csv)
python3 pep_length.py > pep_length.csv
  1. Fuse all the things we've calculated in to one file for analysis
python combine_data.py > dd_stats_complete.csv
  1. Generates a list of all the peptides with a mean distance less than the length of the peptide (good peptides).Can be used to generate the list of bad peptides by changing the angle bracket. (Takes dd_stats_complete.csv as the input file, can output good_pep.csv)
python3 compare_stats.py > good_pep.csv
python3 compare_stats.py > bad_pep.csv
  1. Plots a histogram for the mean distance of the 20 ranks of a specific peptide. Specify the peptide name and path to the directory in the input. (Input: peptide directory, output: graph)
python3 plot_mean.py
  1. Plots a histogram of the mean distance for all the 'bad peptides'. Change the input csv file to plot for the 'good peptides'. (Input:bad_pep.csv/good_pep.csv, Output:graph)(histogram generated is messy and random)
python3 plot_all_means.py

Results as stored in group project space

Project space directory guide -:

peptides:sdf files of the 8000 peptides

fin_results: results output of the final diffdock run for the 8000 peptides

csvs: dd_fin- list of confidence scores for 8000 peptides,
dd_test,dd_test2- inference_steps and samples_per_complex pairs along with confidence score of 3,20 peptides (grid search),
dd_stats2- Mean distance, standard deviation and confidence scores of 8000 peptides
pep_length- Maximum length of the 8000 peptides
bad_pep- list of 'bad' peptides with mean dist and confidence score, good_pep- list of 'good' peptides with mean dist and confidence score

pep_graphs: svg files of histograms of the mean distance for GAA, GSN, LMC and GWC + word file for the histograms of all peptides, bad peptides and good peptides
pep_pymol: pymol files for GAA, GSN, LMC and GWC
(GAA- mean distance lower than the peptide length (good peptide) and narrow standard deviation.
GSN- mean distance shorter than the peptide length (good peptide) but broad standard deviation.
LMC- mean distance longer than the peptide length (bad peptide) and broad standard deviation.
GWC- mean distance longer than the peptide length (bad peptide) but narrow standard deviation.)