Benchmarking computational methods for identifying and quantifying polyadenylation sites from 3′ tag-based single-cell RNA-seq data

This repository contains R scripts that were used to generate results and plots for the paper titled "Benchmarking computational methods for identifying and quantifying polyadenylation sites from 3′ tag-based single-cell RNA-seq data".

1. Data/: This directory contains the reference pA datasets.

2. Rscript/: This directory contains code for running the APA analysis tools.

3. DataProcessing/: This directory contains scripts for data pre-processing of the benchmark study.

functions.R: Common functions called by other scripts.

readPACs.R: This script is used to reads the outputs from APA analysis methods and stores them as a PACdataset object from movAPA for filtering, statistics, and other operations.

sata_pa.R: This script is used to integrate PACdataset objects from different methods for the same sample data and store them as a list for subsequent unified data loading and calculation. Meanwhile, it preprocesses the reference pA dataset.

4. Benchmark/: This directory contains scripts to evaluate the performance of the ten APA analysis methods under various aspects.

identification_benchmark.R: This script is used to compare the pAs identified by different methods and generate the corresponding figures for the manuscript.

uniquePA_benchmark.R: This script is used to compare the consistency of pA identification results across different methods and generate the corresponding figures for the manuscript.

quantification_benchmark.R: This script is used to evaluate the performance of different methods for pA quantification and generate the corresponding figures for the manuscript.

DEAPA_benchmark.R: This script is used to evaluate the performance of different methods for detecting DEAPA genes and generate the corresponding figures for the manuscript.

summary_benchmark.R: This script is used to compare the time and memory consumption of different methods, summarize evaluation results, and generate corresponding figures for the manuscript.