RNA seq data analysis without control samples

Question

[vasanthmandla]

Hi All,

I have RNAseq data (PE) of 37 human cancer samples. The aim of the study is to find novel and transcript isoforms. I have followed the following process:

alignment to reference genome (used reference gtf file while building index) was done using Hisat2, transcript assembly using stringtie2, merged the transcript assembly of all samples using stringtie --merge, used the above transcriptome as reference to re-run the stringtie on all the samples.

I would like to know if I am going in the right direction and if am, which steps should I follow for further analysis.

I know I can use tools such as spliceR or IsoformSwitchAnalyzeR for downstream analysis but I do not have any control samples for comparison (no conditions).

Please, help me on how I can proceed forward with this data. Any kind of help is much appreciated.

Thanks in advance.

Answer 1

[Istvan Albert]

When you attempt a discovery it is against some sort of prior knowledge.

If you are discovering new transcripts your "control" will be the know information up to that point, the best-published reference dataset.

So you do have "control" it is just that other people collected it. Now the job is to validate the new insights and establish if what you found are indeed "new transcripts" and not artifacts of the process.

[vasanthmandla]

Hi Istvan Albert or Anyone,

Thank you for your time and response.

Could you please, expand more on this or suggest any publication that could help me in understanding the methodology for this kind of analysis.

I am sorry for being so naive, I am new to this kind of analysis.

My data belongs to colorectal cancer.