AnalyteExplorer

The goal of AnalyteExplorer is to pre-process data for the AnalyteExplorer module in ImmuneSpace.

Installation

You can install the development version of AnalyteExplorer from GitHub with:

# install.packages("remotes")remotes::install_github("RGLab/AnalyteExplorer")

Workflow with ImmuneSpace

library(AnalyteExplorer)
library(UpdateAnno)
options(debug_dir= tempdir())
labkey.url.base<-"https://datatools.immunespace.org"labkey.url.path<-"/AnalyteExplorer"genes<- process_data("genes")
validate(genes)
res<- update_table(genes)
cohorts<- process_data("cohorts")
validate(cohorts)
res<- update_table(cohorts)
btm<- process_data("blood_transcription_modules")
validate(btm)
res<- update_table(btm)
signatures<- process_data("gene_signatures")
validata(signautres)
res<- update_table(signatures)
summaries<- process_data("gene_expression_summaries")
validate(summaries)
res<- update_table(summaries)

Data Processing

process_blood_transcription_modules()

• Based on Molecular signatures of antibody responses derived from a systems biological study of 5 human vaccines
• The function updates the gene symbols with Hugo.
• This table is used to display metadata about the selected module in the app.

id	name	genes	matched_gene_ontology_terms	number_of_genes	module_category
M0	targets of FOSL1/2 (M0)	CCL2, COL1A2, DCN, IL6, CXCL8, LIF, MGP, MMP1, MMP2, MMP9, PLAU, THBD	extracellular space (11), extracellular region (11), protein binding (9)	12	TF targets
M1.0	integrin cell surface interactions (I) (M1.0)	COL1A1, COL1A2, COL5A1, DPYSL3, MYH10, NRP1, PTK2, RHOC, RRAS, SEMA6A	protein binding (26), axon guidance (20), extracellular region (16)	29	molecular function
M1.1	integrin cell surface interactions (II) (M1.1)	AHSP, ALAD, ALAS2, CPOX, E2F2, FECH, GATA1, HEMGN, HMBS, PLEK2, TMOD1	protein binding (12), extracellular region (12), extracellular matrix structural constituent (9)	12	molecular function
M2.0	extracellular matrix (I) (M2.0)	CD1D, HLA-DMA, HLA-DMB, HLA-DPA1, HLA-DPB1, HLA-DQA2, METTL7A, WDFY4	protein binding (27), extracellular region (26), extracellular matrix (21)	30	location

process_gene_expression_summaries()

• This function creates a gene expression table by cohort, timepoint, and analyte type (gene, blood transcript module, or gene signature).
• Processing steps:
  1. Fetch gene expression matrices and metadata from ImmuneSpace and combine them into one ExpressionSet object.
     • Remove genes that are not available in all expression matrices.
  2. Remove samples that have negative timepoint and select one timepoint if sample has multiple baseline timepoints.
  3. Create gene expression table summarized by analyte type.
     1. In sample level, when summarizing by blood transcript module or gene signature, compute geometric mean of the expression values of the genes in the module or signature
     2. In cohort level, compute the fold change of the expression values for all combinations of timepoints comparing to the baseline timepoint.
     3. Compute mean and standard deviation of those fold change values by analyte type
  4. Merge the three summarized tables

cohort	sample_type	study_accession	condition	timepoint	analyte_id	analyte_type	mean_fold_change	sd_fold_change	id
healthy aldults	Whole blood	SDY1529	Yellow_Fever	0	A1CF	gene	0	0	1
healthy aldults	Whole blood	SDY1529	Yellow_Fever	0	A2M	gene	0	0	2
healthy aldults	Whole blood	SDY1529	Yellow_Fever	3	M0	blood transcription module	-0.0385090	0.0875805	3894315
healthy aldults	Whole blood	SDY1529	Yellow_Fever	3	M1.0	blood transcription module	-0.0181510	0.1051648	3894316
healthy aldults	Whole blood	SDY1529	Yellow_Fever	3	21357945_1_8	gene signature	0.7102219	0.5402085	4028003
healthy aldults	Whole blood	SDY1529	Yellow_Fever	3	21357945_2_9	gene signature	0.1105955	0.5026520	4028004