Mikado code details are available on github Mikado available via bioconda However I used the singularity package created by the lead develepor of BIND pipeline.
ml singularity
singularity pull --name mikado.sif shub://aseetharam/mikado
Consolidate all the transcripts, and predict potential protein coding sequence by Mikado:
-
Step 1: Make a configure file and prepare transcripts
- Step 1.1 Prepare a tab-delimited file called list.txt (shown below) to include gtf path (1st column), gtf abbrev (2nd column), stranded-specific or not (3rd column):
/path/to/file/genome_class.gtf araip_class False /path/to/file/genome_transcripts.gtf araip_cufflinks False /path/to/file/genome_stringtie.gtf araip_stringtie False /path/to/file/genome_strawberry.gtf araip_strawberry False-
Step 1.2 Move or generate a symoblic link to the portcullis_all.junctions.bed file and the genome file
-
Step 1.3 Run step1.sh script
./step1.sh <genome.fasta>-
Step 1.4 Edit the configure.yaml file created from running Mikado_step1.sh See example_configuration.yaml
Edit configure.yaml manually to keep all ORFs. Mikado nosplit mode is selected in step1 and it is best to keep all ORFs if any ORFs overlapped. Add these lines under 'subloci_out:' in configure.yaml
output_format: report_all_orfs: true
-
Step 2: Generate mikado_prepared.fasta Run step2.sh; this script will generate mikado_prepared.fasta file that will be used for predicting ORFs in the next step.
-
Step 3: Predict potential CDS from transcripts There are multiple ways to conduct this step. The output needed from this step is an ORF.bed file. Scripts available to use for this step:
-
Step 4:Pick best transcripts for each locus and annotate them as gene
./step4.sh <ORFs.bed> <prefix>
This script will generate a number of output files. The main output of interest will be labeled as prefix.loci.gff3
It is best to filter the mikado output (prefix.loci.gff3) before going to the next step. Please review filter_output.md.