Tool for Unbiased Literature Searching and Gene List Curation

Designed for genomic and proteomic data analysis, enabling unbiased PubMed searching, protein interaction network visualization, and comprehensive data summarization. This package aims to help users identify novel targets within their data sets based on protein network interactions and publication precedence of target's association with research context based on literature precedence. Methods in this package are described in detail in: Douglas et al. (2025) . Key functionalities of this package also leverage methodologies from previous works, such as: - Szklarczyk et al. (2023) - Winter (2017) .

DeSciDe: A Tool for unbiased gene selection by simple visual analysis.

DeSciDe (Deciphering Scientific Discoveries) is an R package designed for genomic and proteomic data analysis, PubMed search, protein interaction network visualization, and comprehensive data summarization.

Features

Perform PubMed searches for a list of genes and terms.
Search the STRING database for protein interactions.
Rank PubMed and STRING search results to identify high confidence and novel hits.
Visualize results using heatmaps and network plots.
Summarize data from multiple sources.

Installation

Before installing the DeSciDe package, make sure you have BiocManager installed:

# Install Bioconductor dependencies
if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("STRINGdb")
BiocManager::install("ComplexHeatmap")

DeSciDe is available on CRAN, and it can be installed with:

# Install DeSciDe via CRAN
install.packages("DeSciDe")

You can install the development version of DeSciDe from GitHub using the devtools package:

# Install devtools if you haven't already
install.packages("devtools")

# Install DeSciDe from GitHub
devtools::install_github("camdouglas/DeSciDe")

Usage

Basic Example

Below is a basic example of how to use DeSciDe to analyze a list of genes and terms:

# Load the package
library(DeSciDe)

# Define your list of genes and terms
genes_list <- c("JUN", "MYC", "HDAC1", "TRIM33")
terms_list <- c("cancer", "romidepsin")

# Run the analysis pipeline
results <- descide(
  genes_list = genes_list, 
  terms_list = terms_list
)

# View PubMed search results
head(results$summary_results)

Parameters

genes_list: A list of gene IDs.
terms_list: A list of search terms.
rank_method: Method to rank results ("weighted" or "total"). Defaults to "weighted".
species: NCBI taxon ID of the species. Defaults to 9606 (Homo sapiens).
network_type: Type of STRING network to use ("full" or "physical"). Defaults to "full".
score_threshold: Minimum score threshold for STRING interactions. Defaults to 400.
threshold_percentage: Percentage threshold for ranking. Defaults to 20%.
export: Logical indicating whether to export the results. Defaults to FALSE.
file_directory: Directory for saving the output files. Defaults to NULL.
export_format: Format for export ("csv", "tsv", "excel"). Defaults to "csv".

Testing

Due to API rate limits and the potentially long duration of some tests, certain tests are skipped on CRAN using the skip_on_cran() function from the testthat package. These tests are designed to run in local environments and CI/CD pipelines where the environment variable NOT_CRAN is set to "true".

To run all tests locally, set the environment variable as follows: