Merge branch 'main' of https://github.com/schosio/CHOLAR

Author: schosio

README.md

@@ -32,53 +32,3 @@ This will generate
 
 
 
-# Summary
-
-RNA-sequencing has found numerous implementations in research, from distinguishing immune cell subtypes 
-to differential gene expression between cancer versus normal tissue types [@Villani2017][@Bao2021].
-Another application of RNA-seq is to identify novel transcripts involved in various biological processes
-(Gupta, Kleinjans and Caiment 2021). The most relevant is context and cell-type-specific non-coding RNAs,
-such as long non-coding RNAs (lncRNAs), which have become a case-point for most transcriptomic studies proving
-their role in regulating gene expression, post-transcriptional regulation, and epigenetic regulation
-(Engreitz et al. 2016, Zhu et al. 2019).
-It is becoming crucial to check the relative expression of lncRNAs in transcriptome-wide studies. Our group
-has developed an automated lncRNA expression pipeline. The only requirement from the user-side is raw data in
-FASTQ format ( Paired-end or Single-end). The user will get a list of known and novel lncRNAs, and differential
-gene expression between condition(s). The pipelines come with a user-friendly GUI, thereby eliminating the need
-for the user to be versed in complex transcriptome analysis and UNIX environment.  The source code is available
-under an open-source licence at https://github.com/schosio/CHOLAR.
-
-# Statement of need
-
-The number of inferences generated from RNA-seq datasets is countless. The software used in the RNA-seq analysis
-pipeline requires a UNIX-based command-line interactive (CLI) environment, with each software executed in succession.
-Installing multiple CLI tools, handling various file formats, and plotting graphs require understanding of Linux and
-R programming language. Moreover, no tool identifies novel lncRNAs from raw transcriptomic data to the best of our
-knowledge. LncRNA identification tools such as `CPAT` (Wang et al. 2013) (and other tools) take either transcript
-sequence (FASTA) or transcript coordinate (BED, GTF) as input and provide a list of predicted lncRNAs.
-To address these issues, we developed CHOLAR which is a tool for characterization of LncRNA from raw reads .
-`CHOLAR` i) identifies novel lncRNAs from raw reads ii) provides a user-friendly GUI interface to make changes
-at every step iii) allows to identify differentially expressed genes and lists known and novel lncRNAs iv) generates
-publication-quality plots such as MA plot, Volcano plot and heatmap.
-
-# Implementation
-
-The `CHOLAR` pipeline is implemented in bash and R, where it first reads the input FASTQ files(s) to check the
-quality of reads using `FastQC` (Fiancette et al. 2021). Bad quality ( < 28 ) reads and adaptors are removed using
-`Trimmomatic` (Bolger, Lohse and Usadel 2014). `HISAT2` performs the mapping of reads on the human reference genome
-(hg38) (Zhang et al. 2021). The SAM files generated from `HISAT2` are converted to BAM, and PCR duplicates are removed
-utilising the `samtools` toolkit (Danecek et al. 2021). 
-The transcript assembly is done using Stringtie, and the resulting GTF files are merged using the merge utility of
-`Stringtie` (Pertea et al. 2015). The merged GTF file is compared against the reference annotation file from GENCODE
-(Frankish et al. 2021) to filter novel transcripts using `GFFCOMPARE`. The coding potential of novel transcripts is
-predicted using CPAT (Wang et al. 2013). The gene counts are calculated using `HTSeq` (Putri et al. 2022), and subsequent
-differential gene expression analysis (DGEA) is done using packages from R statistical language. 
-The `DESeq2` package is used for performing DGEA (Love, Huber and Anders 2014), and `ggplot` and `dplyr` libraries for
-plotting graphs. The Graphical user interface is built using zenity in bash. The schematic of the tool is given in figure 1.
-
-# Example
-
-We chose the GSE147761 dataset from the GEO database (NCBI) to showcase the CHOLAR tool. A sample of results and plots 
-generated by the tool are given in figure 2. The GUI of the tool is shown in figure 3.
-
-