search

RobertsLab / resources

https://robertslab.github.io/resources/
19 stars 11 forks source link

Extract RNA from Salmon samples #1410

Closed sr320 closed 2 years ago

sr320 commented 2 years ago

It would be good to get these extracted in April. @kubu4 can you tackle the brains, and @mattgeorgephd can you take the other 3 tissues?

We can have Delaney provide support (tube labelling etc)

kubu4 commented 2 years ago

Is there a sample list or anything?

sr320 commented 2 years ago

see https://github.com/RobertsLab/resources/issues/1307 for sample list etc...

kubu4 commented 2 years ago

Brain samples are done. Here's summary table:

They're stored in Sam's RNA Box #3.

Sample Concentration (ng/uL) Elution Volume (uL) Yield (ng)
6A 116 15 1740
7A 132 15 1980
8A 120 15 1800
9A 130 15 1950
10A 94 15 1410
11A 114 15 1710
12A 80 15 1200
13A 80.8 15 1212
14A 74.6 15 1119
15A 108 15 1620
17A 138 15 2070
18A 89 15 1335
19A 67.6 15 1014
21A 61.6 15 924
22A 108 15 1620
24A 83.8 15 1257
26A 98.8 15 1482
27A 110 15 1650
28A 118 15 1770
29A 124 15 1860
30A 102 15 1530
mattgeorgephd commented 2 years ago

@mattgeorgephd can you take the other 3 tissues?

I'll work on these next week. @kubu4 to maintain consistency, which extraction method did you use?

kubu4 commented 2 years ago
mattgeorgephd commented 2 years ago

@kubu4 To confirm - only two tissues remain: liver and gonad?

kubu4 commented 2 years ago

I think that's correct.

mattgeorgephd commented 2 years ago

Salmon samples shipped to UT Austin GSAF on 5/24. Received 5/25. Assigned Job number JA22192.

  1. Sample list and plate map available here
  2. Quote available here
  3. GSAF sample manifest available here
mattgeorgephd commented 2 years ago

@sr320

Received word from the UT Austin sequencing facility regarding the salmon samples. They ran a random selection of samples through their bioanalyzer for a quick QC and forwarded me the results.

Looking at the RIN and curves:

  1. 3/5 brain didn't pass QC.
  2. 1/4 liver didn't pass QC
  3. 0/2 gonad didn't pass QC

Options going forward:

  1. Stick with the original plan and sequence all samples at $75 a pop (omitting ones that failed as mentioned above)
  2. Run bioanalyzer on the remaining samples at $20 a sample, then sequence the ones that pass at $75 a pop

Let me know what you'd like to do

sr320 commented 2 years ago

@mattgeorgephd @kubu4 did y'all run bioanalyzer?

@mattgeorgephd what did your QC look like for your tag-see?

kubu4 commented 2 years ago

did y'all run bioanalyzer

No

kubu4 commented 2 years ago

@mattgeorgephd - Out of curiosity, how do you know which samples are which on those electropherograms?

And, what are their criteria to identify samples that "fail" QC?

mattgeorgephd commented 2 years ago

@sr320 I ran the bioanalyzer on 3-5 samples from each extraction date for the liver and gonad samples to look for consistency. All had two peaks with an RIN above 7.

@kubu4 The GSAF recommends RINs above ~6 for Tag-seq. They said they can compensate for/clean up alcohol contamination (extended tails on the second peak). The well positions are listed above the graphs. Here is a key:

Brain: A1, E1, B2, D2, C3 Liver: D4, E5, F6, G7 Gonad: H8, D11

A third option is to run the bioanalyzer on all of the extra RNA we have in the freezer (most, but not all had extra) and let them know which to run.

sr320 commented 2 years ago

Is it possible samples degraded in transit?

kubu4 commented 2 years ago

@mattgeorgephd - It looks like there are two sets of graphs that are both labelled A1, with drastically different scores. Am I lookin at those correctly?

This gets back to the issue of the impact of whether or not RNA degradation (and, what degree of degradation) has any actual impact on sequencing data. Even if RNA is degraded, if (for example), one was only going to sequence 50bp, then I feel like RNA would have to be severely degraded in order to prevent one from getting usable sequence data.

So, with that in mind, does RNA degradation even matter for sequencing like this? I mean, there's always going to be some level of degradation, even in samples with a RIN of 10. So, we have to decide what level is acceptable/unacceptable...

mattgeorgephd commented 2 years ago

@sr320 There are samples with an RIN of around 10, so degradation due to shipping seems unlikely. The box I sent was oversized for the application just in case it got caught up, which it didn't.

@kubu4 Good catch, it looks like they ran A1 twice. One graph looks flat, so maybe they didn't load enough in one well of the bioanalyzer chip?

re: quality required for Tag-seq - there isn't much in the literature to go off of, and it looks like most references are copy/pasting the quality recommendations for RNA-seq. The argument about Tag-seq being less sensitive to degradation makes sense to me. I just wanted to get your guys feedback before I gave the go ahead, especially given that Sam had mentioned that he had encountered difficulties with the brain extractions.