This repo contains the source code to reproduce the analysis and figures of the paper:

Identification of Clonal Hematopoiesis Driver Mutations through In Silico Saturation Mutagenesis Santiago Demajo, Joan Enric Ramis-Zaldivar, Ferran Muinos, Miguel L Grau, Maria Andrianova, Nuria Lopez-Bigas, Abel Gonzalez-Perez; DOI: https://doi.org/10.1158/2159-8290.CD-23-1416

- BoostDM_ch_analyses
	| 1-Observed_mutations 
	| 2-Blueprints  
	| 3-Discovery_index
	| 4-Benchmark
	|   | - Other_cohort
   	| 5-Experimental_assays
	|   | - DNMT3A
	|   | - TP53
    	| 6-BoostDM_cancer_comparison
	|   | - scripts
    	| 7-Expert_curated_rules_comparison
    	| 8-UKB_analyses
	|   | 0_Clinical_phenotypes
	|   | 1_Post_processing_calling
	|   |   | - polN
	|   | 2_Mutation_overview
	|   | 3_Clinical_associations
	|   | 4_BoostDM_scores_tiers
	|   | 5_Rules_comparison
	|   | 6-Fitness_analysis
	|   | 7-New_exclusive_boostDM
-UKB_variant_calling
	| - part1
	| - part2
	| - refs
	| - U2AF1
- Paper_data
	| - BoostDM-cancer
	| - BoostDM-CH
	|   | - discovery	
	|   | - evaluation/CH
    |   | - model_selection
	|   | - prediction
	|   | - splitcv_meta/CH
	|   | - training_meta/CH
	|   | - Results_crossvalidation_50_iterations
	| - Experimental_data
	|   | - DNMT3A
	|   | - TP53
	| - Expert_curated_rules
	| - Intogen-Cancer
	| - Intogen-CH
	| - pfam

• BoostDM_ch_analyses, Contains all code to reproduce all analyses and figures
  • 1-Observed_mutations, Observed mutations from the three training cohorts
    • Observed_mutations.ipynb
      • Figure 1B
      • Supplementary Figure S3
      • Supplementary Figure S8B
  • 2-Blueprints,
    • Blueprint_models.ipynb
      • Figure 2A-B
      • Figure 2C
      • Figure 1C
  • 3-Discovery_index,
    • Discovery_index.ipynb
      • Supplementary Figure S1
  • 4-Benchmark, Cross-validations...
    • Cross-validation.ipynb
      • Figure 1C
      • Supplemental Figure S1G
    • Fscore50_plots.ipynb
      • Figure 1D
      • Figure 1B
      • Supplemental Figure S2A
    • Precision-recall.ipynb
      • Figure 1E
      • Supplemental Figure S2B
    • Other_cohort, Benchmarking with other cohorts
      • GaoExclusive_benchmark.ipynb
        • Supplemental Figure S3C
      • JapaneseBiobank_benchmark.ipynb
        • Supplemental Figure S3C
    • Results_crossvalidation_50_iterations,
  • 5-Experimental_assays, Benchmarking with experimental assays
    • DNMT3A,
      • DNMT3A_training_cohort.ipynb
        • Supplemental Figure S3A (left)
        • Figure 1F
      • DNMT3A_JapaneseBiobank.ipynb - Supplemental Figure S3A (right)
    • TP53,
      • TP53_Giacomelli_JapaneseBiobank.ipynb
        • Supplemental Figure S3B
      • TP53_Kato_Giacomelli_Training_cohort.ipynb
        • Supplemental Figure S3B
      • TP53_Kato_JapaneseBiobank.ipynb
        • Supplemental Figure S3B
      • TP53_Kotler_JapaneseBiobank.ipynb
        • Supplemental Figure S3B
      • TP53_Kotler_Training_cohort.ipynb
        • Supplemental Figure S3B
      • TP53_Ursu_JapaneseBiobank.ipynb
        • Supplemental Figure S3B
      • TP53_Ursu_Training_cohort.ipynb
        • Supplemental Figure S3B
  • 6-BoostDM_cancer_comparison,
    • BoostDM-CH_cancer_comparison.ipynb
      • Figure 3
      • Supplemental Figure S7
    • scripts, Contain dependencies and functions needed for the plots in the folder
  • 7-Expert_curated_rules_comparison,
    • Expert_curated_rules.ipynb
      • Supplemental Figure S7
      • Supplemental Figure S20C
    • False_positive.ipynb, False positive mutations from Bick et. al.
  • 8-UKB_analyses,
    • 0_Clinical_phenotypes,
      • cancerdata_manually.ipynb
      • cancerdata_variables.ipynb
      • covariates.ipynb
      • cvddata_variables.ipynb
      • Infectiondata_variables.ipynb
    • 1_Post_processing_calling,
      • Postprocessing.ipynb
      • Postprocessing_U2AF1.ipynb
      • Generate_final_files.ipynb
      • FINAL_merge.ipynb
      • polN, Generate pool of normals and excluding list. - poN.ipynb - poN_U2AF1.ipynb
    • 2_Mutation_overview,
      • Mut_overview.ipynb
        • Figure 4B
        • Supplemental Figure S8A
      • MutSig.ipynb
        • Supplemental Figure S8B
    • 3_Clinical_associations,
      • Mut_overview.ipynb
        • Figure 4C-D
        • Figure 5
        • Supplemental Figure S12
        • Supplemental Figure S13A-E
        • Supplemental Figure S14
        • Supplemental Figure S15
        • Supplemental Figure S16
      • Clinical_associations_non_observed.ipynb
        • Supplemental Figure S10A-B left
      • Clinical_associations_one_observed.ipynb
        • Supplemental Figure S10A-B right
      • Ethnicity_associations.ipynb
        • Supplemental Figure S10C
      • Age_association_SNV.ipynb, Age associaton per SNV CH mutation
    • 4_BoostDM_scores_tiers,
      • BoostDM_score_tiers.ipynb
        • Supplemental Figure S9
        • Supplemental Figure S13F
    • 5_Rules_comparison,
      • BoostDM_Rules_comparisons.ipynb
        • Supplemental Figure S18
      • UKB_mut_BoostDM_Rules.ipynb
        • Supplemental Figure S17
        • Supplemental Figure S21A
    • 6-Fitness_analysis,
      • 1_variants_distribution.ipynb, Get limit of detection of the cohort per gene
      • 2_ML_fitness_R882H.ipynb, DNMT3A R882H hotspot fitness calculation
      • 3_All_mutations_fitness.ipynb
        • Supplemental Figure S11
    • 7-New_exclusive_boostDM,
      • BoostDM_Rules_comparisons_overview.ipynb
        • Supplemental Figure S19A-C
        • Supplemental Figure S20A
        • Figure 6A
      • Methods_analysis_exclusive.ipynb
        • Supplemental Figure S19D
        • Supplemental Figure S20E
        • Figure 6E
      • Methods_analysis_intersect_rules.ipynb
        • Supplemental Figure S20B, D, F
        • Figure 6B-D
      • BoostDM-CH_ALL_CHEK2_tests_VAF.ipynb
        • Supplemental Figure S21B
      • BoostDM-CH_ALL_CHEK2_tests.ipynb
        • Supplemental Figure S21c-D

Due to UK Biobank data access policies, UK Biobank data cannot be stored or shared in this repository. Consequently, we are including only the code for the variant calling steps and the generation of figures based on UK Biobank data without the list of variants or other clinical data. For further information about raw UKB data, please contact us.

You can access to boostDM-CH source code and documentation in the boostDM-CH pipeline repository.
You can explore and download the main outputs of boostDM-CH in the boostDM-CH website.