pyPheWAS

PyPheWAS: Phenome-Disease Association Analysis Toolkit for EMR Data

PyPheWAS performs large-scale Phenome-Disease Association Studies (PheDAS) using electronic medical record (EMR) data by mapping diseases and phenotypes from ICD-9, ICD-10, and CPT codes to identify statistically significant associations.

Key Features:

• Data Preparation: Implements cohort censoring and age-matching to ensure comparability and analytical validity.
• Traditional PheDAS Analysis: Conducts association analyses using ICD-9 and ICD-10 billing codes to evaluate disease-phenotype relationships.
• Phenotype Mapping with CPT Codes: Integrates current procedural terminology (CPT) codes to extend phenotype definitions beyond ICD-based mappings.
• Novelty Analysis: Assesses significant disease-phenotype associations to distinguish known findings from potentially novel associations.
• Scalability: Processes EMR cohorts exceeding 100,000 patients within hours for high-throughput analysis.

Scientific Applications:

• Disease-Phenotype Association Studies: Identifies known and novel associations in conditions including Down Syndrome and intellectual developmental disabilities using diverse experimental designs and disease groupings.
• Biomedical Research: Supports investigation of complex genetic and phenotypic interactions through large-scale EMR-based association analysis.

Methodology:

EMR data undergo cohort censoring and age-matching, followed by phenome-disease association analysis through statistical testing of mappings derived from ICD-9, ICD-10, and CPT codes. Significant associations are subsequently evaluated using novelty analysis to determine prior evidence in existing literature.