Publications by Charleen Adams
Simulated Marginal Structural Model (MSM) with Inverse Probability Weighting and Lagged Proteins for Longitudinal Protein Analysis of Incident T2D
1 Marginal Structural Models (MSMs) MSMs estimate the causal effect of time-varying exposures (e.g., protein levels) on outcomes (e.g., T2D). Standard regression models can introduce bias due to: Time-varying confounding – Past protein levels influence future levels and T2D risk. MSMs can use Inverse Probability Weighting (IPW) to adjust for the...
4581 sym R (7861 sym/11 pcs) 1 tbl
xQTLbiolinks hack for PCSK9 and METSIM metabolites
I partially hacked xQTLbiolinks—Jeremy’s suggestion for retrieving GTEx eQTLs and running eQTL-pQTL coloc—to run HyPrColoc on our cis-regions, linking PCSK9 pQTL and METSIM metQTL data. It’s a hack because xQTLbiolinks is designed for one GWAS trait (e.g., one pQTL) with one or more eQTLs, not multiple GWAS traits. While it correctly identi...
687 sym R (19756 sym/71 pcs)
HyPrColoc vignette (by Foley and Staley)
1 Introduction HyPrColoc is Bayesian divisive clustering algorithm for identifying clusters of traits which colocalize at distinct causal variants in a genomic region. The default algorithm can identify clusters of putatively colocalized traits within a vast collection of traits, e.g. 1000s, quickly. For each set of putatively colocalized traits, ...
39399 sym R (22766 sym/28 pcs) 2 img
Colocalization Across Clinical and Molecular Phenotypes
1 Data 1.1 UKB-PPP Proteins (2940 Summary Statistics) Inflammation: 736 Cardiometabolic: 736 Oncology: 735 Neurology: 733 Tally UKB-PPP Categories #!/bin/sh # Directory containing the files dir_path="/Users/charleenadams/ukbppp" # Temporary file for storing category counts temp_counts=$(mktemp) # Loop through all .tar files in the directo...
1442 sym Python (11355 sym/7 pcs)
LDSC
1 Syllables Syllables An🔸ti🔸dis🔸es🔸tab🔸lish🔸men🔸tar🔸i🔸an🔸is🔸m 2 Post-GWAS Method Summary statistics from GWAS are the data source. Mini-Me Post-GWAS Ingests GWAS summary statistics as its data Like two-sample Mendelian randomization (MR),but uses the pattern of association between SNPs Implemented originally i...
3546 sym 1 img
Testing UserGWAS/GenomicSEM for Value with Three Moderately Genetically Correlated Chemokines
1 Bioinformatics Processing: Fetching rsID and Pvals #!/bin/bash # Directories and variables TAR_DIR="/Users/charleenadams/ukbppp/chemokine_list" MAP_DIR="/Users/charleenadams/temp_BI/chemokine_rgs_olink/maps" OUTPUT_BASE_DIR="/Users/charleenadams/temp_BI/chemokine_rgs_olink/processed_ukbppp_chemokine_list" MERGED_DIR="$OUTPUT_BASE_DIR/merged_file...
1315 sym R (27079 sym/30 pcs) 3 img
Mendelian Randomization of METSIM N-acetylalanine (ACY1 Region) and BMI
1 Instrument Selection # Set working directory setwd("/Users/charleenadams/metsim_nacy1_bmi") # Load ACY1 data acy1 <- fread("phenocode-C1110.tsv") # Retrieve TSS data for ACY1 using Ensembl #ensembl <- useMart("ensembl", dataset = "hsapiens_gene_ensembl") ensembl <- useEnsembl(biomart = "ensembl", dataset = "hsapiens_gene_ensembl", mirror = "use...
70 sym Python (5360 sym/6 pcs) 1 img
Latent Factor Multivariate `userGWAS` of Selected Proteins
1 Multivariate GWAS In this analysis, we employ a genomic structural equation model (GenomicSEM) to define a new phenotype: a latent factor that influences 12 proteins we know are influenced by a common SNP. 2 Select Proteins Related to Functional Study # Ensure the destination directory exists mkdir -p /Users/charleenadams/ukbppp/selected_set #...
9914 sym R (33832 sym/17 pcs) 2 img 2 tbl
Fetch 22:30529631:C:A:imp:v1_RPS3AP51_rs2412973 from UKB-PPP
1 Can’t Use Sun Supplementary Table 9 (ST9), BUT Can Extract From Primary Data Sun only provided top SNPs for their most significant results in their Supplemental Table 9 (ST9), not for those of varying P thresholds. But I downloaded the all the primary pQTL files for the EUR discovery set (n = 2940). These are about 1.5 TB compressed (find /User...
966 sym R (4824 sym/6 pcs)
Multivariate `userGWAS` of Four Selected Chemokines: CCL5, CXCL5, CCL8, & CCL7
0.1 Multivariate GWAS In this analysis, we employ a genomic structural equation model (GenomicSEM) to define a new phenotype: a latent factor that influences four CCL5, CXCL5, CCL8, & CCL7. In essence, while the chemokines themselves are directly measurable phenotypes, the latent factor we infer captures the shared genetic influence among these che...
6262 sym R (12663 sym/4 pcs) 2 img 1 tbl