Hello Everyone,
So, unfortunately, I sent my RNA samples to a genome sequencing company; 3 treated and 3 controls, and only 2 treated and 1 control have passed the quality check at their facility.
Also unfortunately, : | , I am going to proceed with the sequencing since this is for my master degree experiment and money and time is a problem. . Any advice? or pep talk haha
Thank youuu
•
written
by
WMA •
1 answer
a correction to your title you have only 1 biological replicate for control (aka no replicate)
most modern tools won't quite run with this setup, but you can go back into the past, deseq1, or tophat those will produce something out of it.
my empirical observation is is that when multiple samples fail there are interconnected reasons and those same reasons probably affect the rest of the data, even those that "pass". Having one replicate by design is usually far more reliable than ending up with one sample because the rest failed.
so make sure you are not spending too much time/effort attempting to salvage something that is not salvageable
Ah interesting, good to know, 1 vs1 did not work once and I assumed the two replicates to be treated independently.
Log in to answer this question.
-
RNA-seq for differential gene expression analysis with only one biological replicates
written by WMA •
-
Drosophila ultra low input RNA seq results - excessive clontech smart cds primers
written by barrbarrbinks •
-
Do I need Illumina Demultiplexing for RNASeq?
written by Hashirama •
-
Could adaptor contamination and primer dimers lead to low number of cells detected during a single …
written by charlesgwellem •
-
Several failures present in QC of RNA-seq data from company via fastqc
written by citronxu •
-
How can I plan my experimental samples to decrease batch effects / allow for better correction?
written by n.tear •
-
RNAseq data analysis help
written by kumars.sv •
-
Regenotype vcf using some samples as reference
written by firestar
-
ERRBS Methylation data analysis
written by madkitty
-
how do I check the sequencing reads
written by Kurban
Not much to comment here. Results are very unreliable, try to focus on genes with larger fold changes and small FDRs, not something like FC=1.24, FDR=0.049. Develop a hypothesis and confirm it with further experiments to be sure that you're not chasing ghosts. It is a rather "poor" experiment to build a thesis on unfortunately to be honest.
Yes, I know :( Thank you.
With confirm I mean try to do experiments that tell you whether the biology the RNA-seq suggests is true. Not validating individual genes, rather pathways or anything you can measure on the cellular level.
I think the only thing to do is on the wet lab side of things - ie if you have genes of interest do RT-qPCR to confirm their differential expression. Unfortunately without multiple replicates the statistics of the RNA-seq experiment won't be super useful, but I think with a little bit of validations experimentally you can still have a great story!!
Thank you all.