transcripts assemble by StringTie

Question

[fatimarasool135]

StringTie can assemble transcripts with or without annotation file. which one is best.?

[fatimarasool135]

I want to find the differential gene expression.

[GenoMax]

Please use ADD COMMENT/ADD REPLY when responding to existing posts to keep threads logically organized. SUBMIT ANSWER is for new answers to original question.

Answer 1

[JC]

Depends on what are you looking for, if you want just to quantify transcript expression in your annotation or if you want to also identify novel transcripts or isoforms. Choose the one that fits your questions.

[fatimarasool135]

Hi JC for differntiall gene expression ...which one I should have to use?

[JC]

Again, if you are fine with the current gene models, use the annotation. Only if you want to know is there are novel genes or isoforms you should use de novo assembly

[fatimarasool135]

how can I do this? I can be possible by de novo assembly by using GtringTie without annotation file.?

[JC]

if you are not using an annotation, you are doing de novo assembly

[fatimarasool135]

I got following error -WARNING: no reference transcripts were found for the genomic sequences where reads were mapped! Please make sure the -G annotation file uses the same naming convention for the genome sequences.

The command is ./stringtie G1_sorted.bam -B -o G1.gtf -G Triticum_aestivum.IWGSC.42.gtf -p 4 -C G1.refs.gtf -A G1.abund.tab Please suggest me how ican resolve it?

[JC]

The error explains by itself, your BAM file used a genome reference with sequence names didn't match your GTF annotation, make sure you are using the same dataset for the genome and annotation, even when the files are named similarly, doesn't mean they differ in chromosome names, like if you reference has chr1, chr2, chr3; your GTF could be indicating 1, 2, 3 as chrom names.

For that plant, you can use Phytozome or Ensembl Plants to get genome and annotation.

[fatimarasool135]

I download the reference genome file from NCBI used for mapping. and annotation file from Ensembl plant. Your mean is that I have to use the annotation file downloaded from NCBI? After that bam file name and annotation file name should be the same?

kindly explain to me again .i did not get.

[JC]

Sadly, even when the genome and version are named equally in NCBI and Ensembl, the sequences names could be different. Use genome and annotation from the same source to ensure you have the same naming.

[fatimarasool135]

Hi JC, the header (name) of sequence from both file downloaded from Ensembl are different.can you suggest me how I can change it. It makes me confusing.

[JC]

are you using: ftp://ftp.ensemblgenomes.org/pub/release-42/plants/fasta/triticum_aestivum/dna/Triticum_aestivum.IWGSC.dna.toplevel.fa.gz ftp://ftp.ensemblgenomes.org/pub/release-42/plants/gtf/triticum_aestivum/Triticum_aestivum.IWGSC.42.gtf.gz ?

[fatimarasool135]

HI JC,

I am following the this paper and i have read many times this paper

Transcript-level expression analysis of RNA-seq experiments with HISAT, StringTie and Ballgown and I have run strigtie sucessfully, after that I have to run the ballgown but could not understand this command.

>pheno_data = read.csv("geuvadis_phenodata.csv")

Acutally I dont know what how to make .csv file of mine samples.what pheno information contain this .csv file.

I have checked the supplementry data of this research article but not found any .csv file.

Please guide me how I can make this .csv file for my samples for step number 8 load the phenotype data for ballgown

Answer 2

[Kristoffer Vitting-Seerup]

Why are you using StringTie in the first place? Not that it is a bad tool but you should only use it if you are interested in specific questions. Take a look at this section of my vignette which describe tools to use for quantifications in different circumstances.

[fatimarasool135]

Thank you.I have no experience of the R language.Are you suggesting me to use Salmon or Kallisto. method?

[Kristoffer Vitting-Seerup]

If you are not interested in any novel transcripts I would suggest using Salmon or Kallisto.

And the R language might come in handy as most tools making it easy to do the downstream analysis are in the form of R packages.

[fatimarasool135]

ok .thank you.I will try this package