In i-shah/httRws: Obtain HTTr data from NCCT webservices

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

httRws

A package to access the EPA/ORD/NCCT high-throughput transcriptomics (HTTr) data.

The httRws package provides functions to search for chemicals that have been run in the HTTr platform and to retreive the raw and processed data on a probe / gene level. These functions utilize the NCCT internal (development) RESTful webservices for accessing HTTr data on ~2,200 chemicals, 8 concentrations, 1 cell type (MCF7), ~58,000 samples, ~21,000 probes, ... Currently, these webservices are only available on the EPA intranet at (httr-dev.epa.gov) and are for internal development and testing purposes only.

Study and treatment

Use searchHTTrChem to find chemicals by searching the name/partial name/regex. Enter a zero space
string will retrieve all chemicals. This function returns a dataframe with all the information needed for subsequent queries. Chemicals are identified by: the blind sample identifier (epa_sample_id) or the dsstox_sid (linked to the chemistry dashboard http://comptox.epa.gov). Currently, the raw count data are retrieved by the epa_sample_id and the differential expression data by the dsstox_sid.

Use getHTTrChemPlates, getHTTrPlateGroups and getHTTrPlateInfo to obtain the complete study design information.

Raw HTTr data

Use getHTTrProbeCounts to get the raw HTTr data in terms of probe counts using the epa_sample_id. Currently, there are ~21,111 probes and the raw count associated with these probes can be obtained at a sample level. Some of the probe have very low probe counts across samples so use the various options for filtering the data (e.g. average count > 10, median count > 50, etc.)

Processed HTTr data

Use getHTTrDEG to get the differentially expressed probes/genes for each chemical by dsstox_sid calculated by DESeq2 by comparing each treatment (chemical,concentration) vs DMSO control counts for each probe across multiple replicates, accounting for plate effects. The default skinny format gives one row per chemical, concentration, probe (gene included), and the log2fc (L2FC) and associated statistics. It is possible to filter these data by p-value, min average probe count (across reps), and |L2FC|. It is also possible to pivot the data into a wide format where rows are chemical treatments and columns are probes/genes.

i-shah/httRws documentation built on May 31, 2019, 5:34 a.m.