Which all tools or packages would you recommend to me for clinical NGS data analysis. The aim is, at the end, I need the output with the best accuracy
Tools or packages which I have chosen
#Trimming the read sequence
Trimgalore and cutadapt or Trimmomatic
#Indexing and read alignment
bwa or bowtie2
#Transcript expression
Tophat2 and cufflinks
#Variant Calling analysis
bwa-picard-gatk pipeline
Mutect2 for [Normal+Tumor]
SomaticSnipper or Mutscan
#Variant Annotation
SnpEff or VarScan
Hi Snijesh
The tools seem to look alright given the data is from Illumina. If not you might want to add some more of them that deal with single end data.
@Garima Ayachit
What if the data are from a different platform? Like SOLiD or Ion Torrent
Ion-Torrent mostly gives processed data due to its in-built tools. It also gives you Variant calls. If you still have to do it outside FastX toolkit, sra-toolkit has some scripts for pre-processing. Cutadapt works with all.
@Garima Ayachit
Are there any other tools or pipeline which can outperform the aforesaid pipeline??
First of all, don't map genomic sequencing data with bowtie, use BWA that'll improve your mapping around indels, this is particularly important.
The listed tools are all over the place, a combination of the abovementioned tools will be ideal depending on what kind of question you are answering. Can you be more specific?
@Yassine Souilmi
First of all, thank you for suggesting BWA over Bowtie.
My plan is to make a tool which detects gene pattern and mutation in a user input sample.
The sample can be on any platform [Illumina, SOLiD, Ion Torrent] so that according to the platform I can design the pipeline yielding the best of best output.
I guess HISAT2 or TopHat2 along with cufflinks can be used to detect the pattern of the genes in the sample (Transcript expression).
What are your suggestions for identifying gene pattern and mutations of genes in the sample?
Thank You
First, what's your data? is it genomic sequencing? RNAseq? I'm not clear hear
That means you need 2 pipelines. The processing for both is fairly different. For genomic DNA you do need a mapper that can work around indels, which means BWA mem. For RNAseq data, you can use bowtie2, but you need to use tophat or similar to solve the splicing question.
@Yassine Souilmi
Ya, I definitely need multiple pipelines. But I want to know, Are there any other tools or pipeline which can outperform the aforesaid pipeline??
For genomic DNA, if you're dealing with human data, just follow GATK best practices. It was validated backwards and forward. And it is the best-supported pipeline by the community. Also, it remains the "golden standard".
For RNASeq, the broad got a draft best practices for RNAseq, but not as well recognised by the community as its DNA counterpart. I would go with something like: bowtie2+STAR or GSNAP for splice-aware mapping, RGASP or GSTRUCT to reconstruct the transcripts, and cufflinks to look at differencial expression. Then the analysis from there goes depending on what you are trying to answer. A good example pipeline is taffeta by Blanca Himes, just look it up on github. it includes several QC steps which are as important as the analysis itself.
@Yassine Souilmi
Thank you for your suggestions I will go through the links
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools