README.md

In pmtran5884/Glycancc: Community Clustering of Anti-Carbohydrate Antibody data

Glycancc

Package to analyze anti-carbohydrate antibody population data

To Install (total time < 5min)

The packages "PCAtools" and "ComplexHeatmap" have to be installed seperately from Bioconductor first before package installation.

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("ComplexHeatmap")

BiocManager::install("PCAtools")

Now install the "Glycancc" package

if (!require("devtools", quietly = TRUE))
    install.packages("devtools")
devtools::install_github("pmtran5884/Glycancc")

Usage (time <5 min)

See "code for figures" folder for detailed code for generating manuscript figures

Load library and set seed

library(Glycancc)

####### set seed ##########
set.seed(123)

Generate K-nearest neighbors network and louvain cluster ACAs

#umap
daisy_umap<-make_umap(t(daisy$igg),n_neighbors = 5,metric = "cosine")

#network
daisy_knn_graph<-knn_network_from_umap(daisy_umap,metric="cosine")

#plot
plot_cluster_graph(daisy_knn_graph, daisy_umap, cluster_method = "louvain")

Linear models of ACA clusters against type 1 diabetes related phenotypes

#calculate the first principal component for each ACA cluster
daisy$eigenmatrix<-make_eigen_matrix(daisy$igg,co_cluster)

#Combine eigenmatrix with daisy phenotype data and add to daisy list and reorder factor levels
daisy$combined<-cbind.data.frame(daisy$eigenmatrix,daisy$pheno)
daisy$combined$Group<-factor(daisy$combined$Group,levels = c("Control","Non-progressor","Progressor"))

#Calculate linear models for each ACA cluster X ~ Group + Sex + FDR + HLA_risk + Draw_Age
daisy_eigen_models<-univariate_models(daisy$combined,
                                      c("Group", 
                                        "Sex", "FDR", 
                                        "HLA_risk","Draw_Age"))

#Convert linear models results to ggplot2 friendly format                                        
daisy_model<-convert_to_ggplotformat(daisy_eigen_models$fulltable)

# forest plot for all ACA clusters
forest_plot_lm_model(daisy_model)

End

For more information, please read the preprint on Research Square "The Anti-carbohydrate antibody repertoire in type 1 diabetes"

Citation

If you use the package and its code, please cite the release from Zenodo (DOI: 10.5281/zenodo.7143430).

pmtran5884/Glycancc documentation built on Oct. 5, 2022, 10:14 a.m.