De novo assemble RNA-seq sequence

·(編集済み)· / , , ·

121

閲覧·

この記事はに最終更新されました。一部の内容が古くなっている可能性があります。ご質問があれば著者にお問い合わせください。

主要な洞察

AI · GEN

Intro

A research paper published on biorxiv determined a new coronavirus subgenus, I would like to figure out is there any changes on protease. However, the sequence data has not been publish.

Fortunately, the similar sequence is do available on NCBI, unfortunately, only RNA-seq data is available.

So I need to assemble the RNA-seq reads first, and BLAST the sequence I need with the assembled data.

TL;DR

Setup the environment with conda:

BASH

conda create -n sra_env -y
conda activate sra_env
conda install -c bioconda -y sra-tools trinity transdecoder blast fastp fastqc

Fetch the data:

BASH

prefetch SRR11301086
fasterq-dump SRR11301086
mkdir -p analysis_results/SRR11301086/{raw_fastqc,clean_fastqc,fastp,trinity,transdecoder,blast_results}

Data quality check

BASH

fastqc  SRR11301086_1.fastq SRR11301086_2.fastq \
        -o analysis_results/SRR11301086/raw_fastqc \
        -t 28 \
        2>&1 | tee analysis_results

Quality control using fastp

BASH

fastp -i SRR11301086_1.fastq \
      -I SRR11301086_2.fastq \
      -o analysis_results/SRR11301086/fastp/SRR11301086_1.clean.fastq \
      -O analysis_results/SRR11301086/fastp/SRR11301086_2.clean.fastq \
      --qualified_quality_phred 20 \
      --length_required 50 \
      --thread 28 \
      --html analysis_results/SRR11301086/fastp/SRR11301086_fastp.html \
      --json analysis_results/SRR11301086/fastp/SRR11301086_fastp.json \
      2>&1 | tee analysis_results/SRR11301086/logs/SRR11301086_fastp.log

Data quality check (post-cleaning data)
CodeBlock Loading...
Assemble with Trinity
CodeBlock Loading...
Check the Trinity result:
CodeBlock Loading...
BLAST sequence of interest
1. Put your sequence in query.fasta.
  CodeBlock Loading...
2. Make BLAST database and run:
  CodeBlock Loading...
Check the BLAST result:
CodeBlock Loading...
Extract the sequence from trinity.Trinity.fasta
CodeBlock Loading...

Tail

You can also blast with the Predicted sequence:
CodeBlock Loading...
Make BLAST database and run:
CodeBlock Loading...

Intro

A research paper published on biorxiv determined a new coronavirus subgenus, I would like to figure out is there any changes on protease. However, the sequence data has not been publish.

Fortunately, the similar sequence is do available on NCBI, unfortunately, only RNA-seq data is available.

So I need to assemble the RNA-seq reads first, and BLAST the sequence I need with the assembled data.

TL;DR

Setup the environment with conda:

BASH

conda create -n sra_env -y
conda activate sra_env
conda install -c bioconda -y sra-tools trinity transdecoder blast fastp fastqc

Fetch the data:

BASH

prefetch SRR11301086
fasterq-dump SRR11301086
mkdir -p analysis_results/SRR11301086/{raw_fastqc,clean_fastqc,fastp,trinity,transdecoder,blast_results}

Data quality check

BASH

fastqc  SRR11301086_1.fastq SRR11301086_2.fastq \
        -o analysis_results/SRR11301086/raw_fastqc \
        -t 28 \
        2>&1 | tee analysis_results

Quality control using fastp

BASH

fastp -i SRR11301086_1.fastq \
      -I SRR11301086_2.fastq \
      -o analysis_results/SRR11301086/fastp/SRR11301086_1.clean.fastq \
      -O analysis_results/SRR11301086/fastp/SRR11301086_2.clean.fastq \
      --qualified_quality_phred 20 \
      --length_required 50 \
      --thread 28 \
      --html analysis_results/SRR11301086/fastp/SRR11301086_fastp.html \
      --json analysis_results/SRR11301086/fastp/SRR11301086_fastp.json \
      2>&1 | tee analysis_results/SRR11301086/logs/SRR11301086_fastp.log

Data quality check (post-cleaning data)

BASH

fastqc analysis_results/SRR11301086/fastp/SRR11301086_1.clean.fastq \
       analysis_results/SRR11301086/fastp/SRR11301086_2.clean.fastq \
       -o analysis_results/SRR11301086/clean_fastqc \
       -t 28 \
       2>&1 | tee analysis_results/SRR11301086/logs/SRR11301086_fastqc_cleaned.log

CodeBlock Loading...

Assemble with Trinity

BASH

nohup Trinity --seqType fq \
        --left analysis_results/SRR11301086/fastp/SRR11301086_1.clean.fastq \
        --right analysis_results/SRR11301086/fastp/SRR11301086_2.clean.fastq \
        --CPU 28 --max_memory 48G \
        --output analysis_results/SRR11301086/trinity \
        2>&1 > analysis_results/SRR11301086/logs/SRR11301086_trinity.log &

mv analysis_results/SRR11301086/trinity.Trinity.fasta* analysis_results/SRR11301086/trinity/

CodeBlock Loading...

Check the Trinity result:

BASH

TrinityStats.pl analysis_results/SRR11301086/trinity/trinity.Trinity.fasta \
2>&1 | tee analysis_results/SRR11301086/logs/SRR11301086_trinity_stats.txt

CodeBlock Loading...

BLAST sequence of interest

Put your sequence in query.fasta.
BASH
vi query.fasta
CodeBlock Loading...

Make BLAST database and run:

BASH

makeblastdb -in analysis_results/SRR11301086/trinity/trinity.Trinity.fasta -dbtype nucl -out analysis_results/SRR11301086/trinity/SRR11301086_nuc_db

tblastn -query query.fasta \
        -db analysis_results/SRR11301086/trinity/SRR11301086_nuc_db \
        -out analysis_results/SRR11301086/blast_results/SRR11301086_nuc_blast.txt \
        -evalue 1e-5 \
        -word_size 2 \
        -outfmt "6 qseqid sseqid pident length mismatch gapopen qstart qend sstart send evalue bitscore qseq sseq" \
        -num_threads 28

CodeBlock Loading...

Check the BLAST result:

BASH

cat analysis_results/SRR11301086/blast_results/SRR11301086_nuc_blast.txt

CodeBlock Loading...

Extract the sequence from trinity.Trinity.fasta

BASH

# Build index
samtools faidx analysis_results/SRR11301086/trinity/trinity.Trinity.fasta
# Extract the sequence
samtools faidx analysis_results/SRR11301086/trinity/trinity.Trinity.fasta TRINITY_DN284_c0_g1_i7 > analysis_results/SRR11301086/blast_results/SRR11301086_TTRINITY_DN284_c0_g1_i7.fasta

CodeBlock Loading...

Tail

You can also blast with the Predicted sequence:

BASH

TransDecoder.LongOrfs -t analysis_results/SRR11301086/trinity/trinity.Trinity.fasta\
    2>&1 | tee analysis_results/SRR11301086/logs/SRR11301086_transdecoder_long.log

TransDecoder.Predict -t analysis_results/SRR11301086/trinity/trinity.Trinity.fasta \
    2>&1 | tee analysis_results/SRR11301086/logs/SRR11301086_transdecoder_predict.log

mv *.transdecoder* analysis_results/SRR11301086/transdecoder/

CodeBlock Loading...

Make BLAST database and run:

BASH

makeblastdb -in analysis_results/SRR11301086/transdecoder/trinity.Trinity.fasta.transdecoder.pep \
            -dbtype prot \
            -out analysis_results/SRR11301086/transdecoder/SRR11301086_db \
            2>&1 | tee analysis_results/SRR11301086/logs/SRR11301086_makeblastdb.log

blastp -query query.fasta \-db analysis_results/SRR11301086/transdecoder/SRR11301086_db \
       -out analysis_results/SRR11301086/blast_results/SRR11301086_prot_blast.txt \
       -evalue 1e-5 \
       -outfmt "6 qseqid sseqid pident length mismatch gapopen qstart qend sstart send evalue bitscore qseq sseq" \
       -num_threads 28 \
       2>&1 | tee analysis_results/SRR11301086/logs/SRR11301086_blast.log

CodeBlock Loading...