README.md

# AlphaFold-NMR

Scripts and data corresponding to Huang, Ramelot, Spaman, Kobayashi, Montelione (2025) "Hidden Structural States of Proteins Revealed by Conformer Selection with AlphaFold-NMR"

<img width="1161" alt="flowChart" src="https://media.github.rpi.edu/user/352/files/b6534f5c-e03d-4529-a4a5-bbc697c9db04">

# Usage

Data and scripts for CDK2AP1-doc1 AF-NMR analysis: 

1. AI enhanced sampling: CDK2AP1-doc1/EnhancedSampling/

  * doc1_noN.fasta: input fasta sequence. We exclude the long disordered tails and non-native tags from the input fasta sequence for AF modeling, to avoid potential influence on the pTM and \<pLDDT\> scores. 
     
  Commands: 

  > sbatch run_doc1_noN.sh (running with slrum) 
  
  This command calculates and relax all 6000 models using run_afsample6000.sh <br>
  The output models are here: AF_models_dropout/doc1_noN. pTM score is reported here: AF_models_dropout/scores.sc and log from AF: slurm-xxx.out 
  
  > python FilterAF2.py -log slurm-xxx.out -rel -inD AF_models_dropout/doc1_noN -outD filteredModels
  
  This command filters out bad models based on the AF log file (e.g. slurm-xxx.out). The python code is copied from here: https://github.rpi.edu/RPIBioinformatics/FilteringAF2_scripts

  Additional processing scripts: 
  
  Merge two chains into one chain for clustering analysis
  > sh runMergedChain.py filteredModels mergedModels
  
  This command finds all pdb file in the fileredModels, merge two chains (using mergeChain.py) and save them in the mergedModels directory. 
  
  5984 models with one merged chain (CDK2AP1-doc1/ESmodels/) are used for the following analyses:  
  
2. Clustering
  
R scripts for CDK2AP1: 

 CDK2AP1-doc1/Clustering/:
    dmPCAClustering.R --> output: pc_dm_pdbs.RData, cluster_pc_dm.csv (in Rstudio, set the working dir to this dir before running the R script) 
 
 We found that "ward methods" gives largest agglomerative coefficient. Number of clusters --> by viusal inspection of "Dendrogram" and pc plots to identify number of well-seperated clusters.  
 
3. Scoring

NMRdata: 
 * CDK2AP1-doc1/RPF/: directory to calculate RPF scores: 
   - Input/
   - 
 * CDK2AP1-doc1/RCI/: directory to calculate RCI and SCC scores
   - RPFtable2SHIFTY.py: convert the chemical shift file used by RPF (bmrbtable file) to SHIFTY format 
   - nmrstar3toSHIFTY-fromBMRB.py: give the bmrb ID number, download chemical shift assignments from the BMRB database and convert to SHIFTY format
   - nmrstar3toSHIFTY.py: convert the local bmrb file in nmrstart 3.0 format to SHIFTY format 
 
 Other scores: 
 * CDK2AP1-doc1/Scoring/: 
  - combine all scores: 
  - plots: 
  
4. State combination
  
scripts for CDK2AP1:
  
5. Validation
  
scripts for CDK2AP1: