bagRNA is a user-friendly ncRNA-aware genome annotation pipeline for eukaryotic organisms. It is designed to predict, correct, classify and assign function to gene models, all in one go!

The pipeline integrates:

• ab initio prediction
• transcript alignment
• homology-based annotation

It identifies both protein-coding and non-coding RNAs (e.g., lncRNA, tRNA, rRNA), inferring biological function and incorporating it into the final annotation.

👤 Main Developer Gabriele Rigano

Bioinformatics and Computational Genomics LAB

University of Messina, Sicily, Italy

📧 gabriele.rigano@unitn.it

📌 Features

• ncRNA-aware: Retains and classifies non-coding features
• All-in-one workflow: From raw data to structural and functional annotation
• Minimal inputs: Genome FASTA, protein evidence, and RNA-seq reads
• Modular & reproducible: Uses Docker containers for tool management

⚙️ Pipeline Overview

1. Data preparation
2. Transcriptome assembly
3. Gene model selection & correction
4. Structural annotation
5. Functional annotation

🚀 Installation

1. Clone the Repository

git clone https://github.com/BCGL-Unime/bagRNA.git

2. Set Up the Conda Environment

conda create -n bagRNA mamba -y
conda activate bagRNA
mamba install -f bagRNA.yaml

3. Download Docker Images

chmod +x download_docker_images.sh
bash download_docker_images.sh

🧩 Database Setup

chmod +x download_databases_bagRNA.sh
bash download_databases_bagRNA.sh

⚠️ Tools to be manually dowloaded by the user

The following must be downloaded manually due to licensing restrictions:

1. GeneMark-ETP
2. Phobius
3. SignalP 6.0 (Download the fast or slow-sequential models)

Running the pipeline

🧪 Real Usage Example

bagRNA.sh \
  --mikado_config mikado_config.tsv --busco_lineage sordariomycetes \
  --Helixer_gff helixer_SS02.fasta.gff --input_fasta maskedSS02.fa \
  --locus_tag SS02 --STAR_manifest mold_reads_for_STAR.tsv yeast_reads_for_STAR.tsv \
  --RAM_limit_Trinity 45G --limitBAMsortRAM 4571838164 \
  --orientation RF --strandedness secondstrand --databases /home/gab/databases_bagRNA/ \
  --max_intron_length 3000 --threads 12 --max_gene_length 30000 \
  --prot_evidence SSref_prot_and_evidence.faa --species "Sporothrix_schenckii" \
  --strain SS02 --codon_table 1 --scoring scerevisiae.yaml \
  --lifted_annotation lifted_ren_SS02.fasta.gff  \
  --submission_template /home/gab/tools/submission_template.sbt --no_tmbed \
  --GeneMark_PATH /home/gab/tools/GeneMark-ETP/ --jaccard_clip \
  --phobius_PATH /home/gab/tools/phobius/

🗂 Required Inputs

Argument	Description	Tips and specifics
--input_fasta	Genome FASTA file (preferably softmasked)	Run EarlGrey or EDTA TE prediction beforehand and get a softmasked assembly; remove mitochondrial/plastidial contigs
--prot_evidence	Protein evidence in FASTA format	Download proteins of your taxon from UniProt
--busco_lineage	BUSCO lineage (e.g., sordariomycetes)	Use the phylogenetically lowest taxon possible
--STAR_manifest	STAR manifest TSV file(s)	Check the example in the repository and the docs
--mikado_config	Mikado configuration tsv file	Check the example and docs, don't change names in column 1
--scoring	Select a Mikado scoring config (e.g., scerevisiae.yaml, plant.yaml, mammalian.yaml)	Check the tutorial here
--submission_template	.sbt file for GenBank submission	Check the example in the repository
--species	Species name in quotes (e.g., "Arabidopsis_thaliana")	Must be written in quotes with underscores instead of spaces
--phobius_PATH	Path to phobius.pl executable	Provide the absolute path of your phobius installation where the "phobius.pl" executable is located /home/gab/tools/phobius/

🧠 Choose one of the two following options

Argument	Description	Tips and specifics
--Helixer_lineage	Choose among fungi, land_plant, vertebrate, invertebrate	Requires prior GPU and NVIDIA Container Toolkit configuration
--Helixer_gff	Use precomputed Helixer GFF	Run Helixer here and provide the GFF as input

📥 Optional Inputs

Argument	Description	Tips and specifics
--lifted_annotation	Liftoff/Lifton GFF annotation	Lift a reference annotation to your genome assembly and provide it as evidence. Lifton works well
--GeneMark_PATH	Path to GeneMarK-ET/ETP executable	Provide the absolute path of your GeneMark installation where the "gmes_petap.pl" executable is located /home/bag/tools/GeneMark-ETP/
--databases	Path to functional annotation databases	Provide the absolute path to where you downloaded the databases with the script "download_databases_bagRNA.sh"

⚙️ Performance / Misc

Argument	Default	Description	Tips and specifics
--threads, -t	1	Number of CPU threads	Number of CPU threads you want to use, try not to use them all, as your system may crash
--jaccard_clip	(on)	Enable for high gene density	Use this flag if you are expecting high gene density with UTR overlap, recommended for fungi
--max_gene_length	30000	Max length of gene models	Check the longest gene in your organism or closely related and set around that length as cut-off
--RAM_limit_Trinity	45G	Max RAM for Trinity	Max RAM for Trinity transcript assembly
--limitBAMsortRAM	4571838164	STAR BAM sort RAM limit	Max RAM for STAR BAM sorting
--orientation	RF	Read orientation (FR, RF, etc.)	Check this tutorial for understanding strandedness
--strandedness	secondstrand	Strand type (e.g., secondstrand)	Check this tutorial for understanding strandedness
--max_intron_length	3000	Max intron size	Check the longest intron in your organism or closely related and set around that length as cut-off
--codon_table	1	NCBI codon table ID	Check the codon tables here
--strain	strain	Strain/isolate name	Assign a strain or isolate name
--locus_tag	bagRNA	Locus tag prefix	Locus tag prefix, here the specifications

🔌 Disable Specific Modules

Argument	Description	Tips and specifics
--no_functional_anno	Skip functional annotation	Skips InterProScan EggNOG KOfamscan, Infernal, Signalp6, Phobius, Tmbed, EffectorP-3.0, AntiSMASH
--no_effectorp3	Skip EffectorP-3.0	Skips apoplastic and cytoplasmic effectors prediction
--no_antismash	Skip AntiSMASH	Skips Secondary Metabolite clusters prediction (fungi only)
--no_tmbed	Skip Tmbed	Tmbed is very fast on GPU but very slow on CPU, use this flag if your machine does not have a performing GPU

How to Cite Work in progress...