# Re-Reproducing differential methylation analyses, again

Having just given a talk on reproducibility, I am in the midst of responding to reviewer comments about what we did (12 months ago!) and boy can I say every minute of putting this notebook together was worth it. I even found where we ran the entire notebook, so all result files are easily accessible. Beyond praising Claire, I will document my follow up analysis here.

Essentially the want more quantitative information on differential methylation beyond ..

Makes sense.

Here is what was originally done.

For example the file named linexon contained 16 exon_intersect_DML_lin_u.txt. The 4 files were concatenated to produce lintable ….

and a little awk

awk 'FNR==NR{sum+=$1;next}; {print$0,sum}' lintable{,} > lin_total
awk '{print $2,$1, $3, (($1/\$3)*100)}' lin_total > lineage_DMLs

to create lineage_DMLs

Analogously here are the developmental_DMLs….

And we certainly need to now how many all_CGs we have…

Table

Feature Family specific DMLs Developmental specific DMLs
Transposable Element 17 16
Promoter Region 2 3
Exon 16 12
Intron 25 46

I know we did this before, but I believe the reviewers want a break-down, or list of which specific transposable elements. This is a long shot if I can find this…
2 minutes later https://github.com/sr320/ipython_nb/blob/master/BiGo_larvae_manuscript4.ipynb.

To be sure files are accurate, I will intersectbed again. Based on recollection there is likely not a difference in proportion based on all TEs. This brings up a an important point of how to record “negative” data that does not go into a paper.

# qPCR – Withering Syndrome ddPCR Comparison

Performed qPCR on the same exact samples previously sent to Alice Nguyen at the Marine Science Institute for digital droplet PCR (ddPCR) to compare our results to hers. These samples were previously quantified as part of the Ab Endo Project (Ab Endo 2011 Water Filter DNA samples). Here’s the list of samples that were sent to (and back from) Alice:

1. CI SRI CP 1A
2. CARMEL +500M 2
3. CI SRI CP 2B
4. CI SRI CP 2A
5. CI SRI CP 1B
6. CARMEL +500M 1

Master mix calcs are here: 20150415 – qPCR WS Ab Endo vs ddPCR

Standard curve was p16RK7 (from 20120730).

Cycling params, plate layout, etc. can be found in the qPCR Report (see Results below).

Results:
qPCR Report (PDF): Sam_2015-04-15 15-38-13_CC009827.pdf
qPCR Data File (CFX96): Sam_2015-04-15 15-38-13_CC009827.pcrd

### Table – Summary of qPCRs and ddPCR data

 Sample Number Sample Name Initial qPCR Copies ddPCR Copies Today qPCR Copies AB01 CI SRI CP 1A 0 3.8 0 AB02 CARMEL +500M 2 1.57 8.7 1.3 AB03 CI SRI CP 2B 147 104 59.2 AB04 CI SRI CP 2A 0 175 83.1 AB05 CI SRI CP 1B 74.8 2.8 0 AB06 CARMEL +500M 1 1.08 4.9 0

The table certainly shows some inconsistency, both between ddPCR and qPCR. Additionally, there is also inconsistency between the initial qPCRs and today’s qPCRs. Due to the limited sample volumes remaining for these sub-samples that were sent to Alice and then returned to us, I will repeat the qPCR using the stock DNA and see how that compares.