Skip to content

Latest commit

 

History

History
115 lines (72 loc) · 4.85 KB

Readme.md

File metadata and controls

115 lines (72 loc) · 4.85 KB

Genome annotation methods

Transposable elements (TEs) annotation

The Extensive de-novo TE Annotator (EDTA) is applied to annotate TEs which includes LTRharvest_parallel,LTR_FINDER_parallel,LRR_retriever,TIR-Learner,HelitronScanner,RepeatModeler,RepeatMasker. This pipeline is integrated in current directory ./workflow/Snakefile and the final TE.gff3 is as input file for following gene annotation (Figure 1).

Protein-coding gene annotation

The gene annotation pipeline:

  • Homology-based prediction

    ​ 83 public RNA-seq dataset including various tissues from BIG, CER, PIM groups are mapped to assembly using hisat2 and subsequently assembled into transcripts by stringtie. Then, TACO is applied to merge all stringtie gtf.

    ​ Proteins from SwissProt Viridiplantae, LA2093, DM v6.1, Ath, ITAG4.0 are integrated and remove redundant proteins by CD-HIT . ESTs are retrieved from the NCBI.

  • Ab initio prediction

    ​ GeneMark-ET/-ES,BRAKER ,SNAP and Augustus are applied to predict gene structure.

  • Integration homology-based and ab initio prediction

​ The integrated gene models with AED values < 0.5 are retained

The TEs and protein-coding gene annotation pipeline is deposited in current directory 1.Repeat_gene_annotaion

To configure this workflow, you can modify config/config.yaml according to needs, following the explanations provided in the file:

  • Add samples to config/samples.tsv. Only the column sample is mandatory, but any additional columns can be added.
  • For each sample, add one or more sequencing units (runs, lanes or replicates) to the unit sheet config/units.tsv. For each unit, define platform, and either one (column fq1) or two (columns fq1, fq2) FASTQ files (these can point to anywhere in your system).

Workflow

image-20220101234140544

Figure 1. Overview of the genome annotation pipeline.

Pooled RNA-seq and assembly model (PRAM) annotation

To discover novel transcripts located in the intergenic region of SL5.0, , which based on on a single step that encompasses all transcript construction, is used. A total of 332 RNA-seq sequences are equipped.

Build SL5.0 genome index by STAR

STAR --runThreadN 20 \
    --runMode genomeGenerate \
    --genomeDir /your_path/STAR/genome.index \
    --genomeSAindexNbases 13
    --genomeFastaFiles /your_path/SL5.0.genome.fa \
    --sjdbGTFfile /your_path/genome.annotation.gtf \
    --sjdbOverhang 149

Map the 332 RNA-seq clean data

STAR --runThreadN 20 --genomeDir /your_path/STAR/genome.index \
--readFilesCommand gunzip -c --outSAMtype BAM SortedByCoordinate \
--outFileNamePrefix sample --outSAMattrIHstart 0 \
--readFilesIn ./clean_fq/{sample}_1.fq.gz  ./clean_fq/{sample}_2.fq.gz

PRAM is used to identify the intergenic region (https://bioconductor.riken.jp/packages/3.9/bioc/vignettes/pram/inst/doc/pram.pdf)

source activate RNA
library("pram")
in_gtf=system.file('extdata/SL5.0.annotation.gtf', package='pram')
bam_data<-read.table("/public/home/zhangzhiyang/anaconda3/envs/RNA/lib/R/library/pram/extdata/399_bam/bam_list",header=F)
pred_out_gtf=tempfile(fileext='.gtf')
pram::runPRAM( in_gtf, in_bamv,pred_out_gtf,method='plst',stringtie='~/anaconda3/envs/RNA/bin/stringtie')

Filter the potential transposable elements, according to information from pram.fasta.rexdb.dom.tsv to remove the potential transposable elements

/public10/home/sci0009/scripts/TEsorter.sh
source /public10/home/sci0009/miniconda3/bin/activate edta
TEsorter pram.fasta   -p 64

Transdecoder is used to predict the likely coding regions in remains

TransDecoder.Predict -t target_transcripts.fasta

Gene function annotation

Homolog annotation of 46 tomato genes by using UniProtKB/SwissProt database

~/anaconda3/bin/diamond makedb --in swissprot.fasta  -d swissprot
~/anaconda3/bin/diamond blastp --more-sensitive -d swissprot -q 46_genome_protein.fasta -k 20 -e 0.00001 -o swissprot.xml -f 5

python xml2tab.py
python UniProt2GO_annotate.py idmapping.tb.gz swissprot_sprot.tab swissprot_sprot.go

The motifs and domains of predicted genes were predicted using InterProScan

source activate interproscan
for i in `less  /public10/home/sci0010/project/17.function/all_prot/MM/MM.pep.fa.split/id`; do echo "/public10/home/sci0009/software/InterProscan/interproscan-5.48-83.0/interproscan.sh    -i /public10/home/sci0010/project/17.function/all_prot/MM/MM.pep.fa.split/${i} -o tsv/${i}  -goterms -iprloo1kup -pa -f TSV -dp -cpu 2 &";done > run.sh
sbatch -p amd_io run.sh
/public10/home/sci0009/software/InterProscan/interproscan-5.48-83.0/interproscan.sh -i  /public10/home/sci0010/project/17.function/all_prot/SL5/SL5.pep.fa.split/SL5.pep.part_001.fa -o tsv/SL5.pep.part_001.fa  -goterms -iprloo1kup -pa -f TSV -dp -cpu 2