Thursday, March 19, 2015

Does having a spec cassette make a difference?

I was asked to check if the strains with the spec cassette have anything unusual in terms of gene expression.

First I needed to determine which samples had the spec cassette. To do this, I ran a grep command on the FASTQ files of may samples and looked for the string "GGGGATCCGTCGACCTGCAGTTCGAAG" which is a sequence in the spec cassette.

gunzip -cd filename.fq.gz | grep GGGGATCCGTCGACCTGCAGTTCGAAG

The following samples had reads that match this string:
cnmx
antx
toxx

The following samples did not have reads that match this string:
taxx
everything else

I checked every sample for strains in red and they were all consistent within a given strain. I only checked a single condition for everything else and not a single match was found in these.


For these marked strains, I pulled out all the genes that were significantly different compared to KW20 under the same condition. (ie. cmnx_M0 vs kw20_M0, antx_M1 vs kw20_M1, etc). Among these 12 samples, there was a total of 222 differentially expressed genes compared to KW20 (190 unique genes)

Results:

Only one gene was found differentially expressed in more than 3 samples. HI_0658 (ATP-binding cassette, subfamily F, member 3) was found differentially expressed in 4 (33%) of the samples, however all four samples were antx so this is not a spec-specific effect.

A few genes (~10 in total) show up in 3 samples. HI_0062 (dnaK suppressor protein, ~3x downregulated) and HI_0235 (alternative ribosome-rescue factor, ~5x downregulation) are differentially expressed compared to KW20 in cmnx_M3, antx_M3 and toxx_M3.

However, looking at another comparison, sxyx_M3 vs KW20_M3 (chosen at random), I see the same behaviour for these two genes. Perhaps it could be due some skewing of the data due to the problematic KW20_M3_C replicate that Rosie mentioned in the previous post.

A operon containing ribosomal protein HI_0776-779 (~2.5x upregulated) is differentially expressed compared to KW20 in cmnx_M1, antx_M2 and toxx_M2.

In summary, if there is any effect caused by having a spec cassette, is it not strong enough to really alter the set of differentially expressed genes.

2 comments:

Rosie Redfield said...

Wait! taxx doesn't have the spc cassette? Isn't that supposed to be the toxin/antitoxin double mutant that Sunita made? I'm sure she said that it is 'marked'. Could it be marked with something other than the usual spc cassette? I'll email Sunita and Josh for clarification.

Rosie Redfield said...

Another comment. The two cmnx samples should have the spc cassette. So should the two toxx samples Anni and I made. But the toxx sample that's part of the original set should be unmarked. So you want to be sure these are separated before doing the analysis.

The one genuine antx sample should be unmarked - it should not have the spc cassette.