Yaamini’s Notebook: Remaining Analyses Part 19

Digital drawing board: Round 2

Steven commented on my previous post and said he liked the heatmap option! In my R script I played around with row clustering and color schemes. There are three decisions that need to be made:

  1. All data vs. differentially expressed subset
  2. Row clustering vs. no row clustering
  3. Color schemes

Decision 1: All data vs. differentially expressed subset


Figure 1. All data, now row clustering, standard color scheme


Figure 2. Subset, no row clustering, standard color scheme

Decision 2: Row clustering vs. no row clustering


Figure 3. All data, row clustering, standard color scheme


Figure 4. Subset, row clustering, standard color scheme

Decision 3: Color schemes
















// Please enable JavaScript to view the comments powered by Disqus.

from yaaminiv.github.io http://ift.tt/2CpnvOS

Yaamini’s Notebook: Remaining Analyses Part 18

My better figure wasn’t actually good

I went back to the (digital) drawing board to make some more options.

Option 1: Dot charts

But this time, larger dots.


Figure 1. Average peptide abundance data across all sites.


Figure 2. Differentially expressed average peptide abundance across all sites.

Option 2: Heatmaps

There aren’t large differences in the colors themselves (especially when looking at the figure with all peptide abundance data), so it’s not that exciting.


Figure 3. Average peptide abundance heatmap.


Figure 4. Differentially expressed average peptide abundance across all sites.

Option 3: Bubble plot

I like Figure 6 more than Figure 5. If possible, I could scale all data by the smallest value to make the bubble sizes bigger and more dramatic.


Figure 5. Average normalized abundance across sites for each peptide. The points themselves are scaled such that larger points represent higher peptide abundance.


Figure 6. Average normalized abundance across sites. Points are scaled such that larger points represent higher peptide abundance, but all site data is grouped together. Points are color coded by peptide.

Option 4: Beeswarm plots

Suggestion from Steven! I think it’s a step up from the dot charts, but maybe not the best.


Figure 7. Beeswarm plot.

I wanted to make a volcano plot but it did not seem possible with my dataset. Perhaps I’m not searching for the right things in Google.

// Please enable JavaScript to view the comments powered by Disqus.

from yaaminiv.github.io http://ift.tt/2oys5kh

Laura’s Notebook: Oly DNA Extraction, testing protocol

Using PAXgene tissue DNIA kit Processing 4 extra samples: 2 @ low pH, 2 @ ambient pH Following protocol, with a couple exceptions

  • Using Tissue Tearer for homogenization
  • Incubation temp/RPM adjustments here

Selected 4 samples from HL, 2 male + 2 female

  • HL-6-18, Female, Cassette #12, tissue mass ~17.2ug
  • HL-6-20, Male, Cassette #12, tissue mass ~11.2ug
  • HL-6-17, Male, Cassette #12, tissue mass ~12.7ug
  • HL-6-10, Female, Cassette #2, tissue mass ~10.5ug

Reviewed slides corresponding to these tissues under scope to identify location of gonad. imageimage

Used razor blad to cut gonad out of paraffin blocks. Identified area with gonad tissue using microscope slides, did my best to only cut out gonad tissue, but this is difficult b/c gonad is such a small area. I’m concerned that I’m unable to be confident in isolating only gonad tissue – it’s likely to be contaminated with gill, mantle, digestive gland, etc. I could pull DNA from the section on on the slide; I wouldn’t get much DNA but perhaps it would be enough and I could be confident in the tissute type, which I presume is very important in methylation analysis, as it likely differs by tissue.

Weighed collection tubes pre- and post- tissue to estimate the tissue mass I extracted from each block. (Goal is 10-20mg tissue, no more than 20mg.)

After removing ethanol supernatant (after xylene, pre TD1 buffer), incubated at room temp for ~10 mins, then 37C for ~15 minutes to evaporate ethanol fully.

Homogenized tissue in TD1 buffer with the Tissue Tearor for ~20 seconds at setting #15.

Opted to NOT add the RNA enzyme to digest RNA; therefore there should be RNA in my samples still.

Both 1-hr incubation periods extended to 1:15, at ~400 rpm (max RPM on Graham Young’s incubator/shaker). NOTE: Graham does have a thermomixer that achevies 1300rpm, but only 70C – I could use this for the 1st incubation, but need to determine whether there is a large enough tube holder – OR instead try using smaller tubes for the incubation steps (if I can).

HICCUP: Sometime during the 2nd incubation (at 80C) the shaker stopped shaking; the incubator likely busted a belt, as the motor sounds like it still works. So, the samples were incubated for 80C for 1:15, with unknown time at 400rpm (at most 45 minutes).

Did not have enought time to perform assay to quantify samples immediately after extration; eluted samples were stored in the refrigerator.

from LabNotebook http://ift.tt/2GvgnOV

Yaamini’s Notebook: Remaining Analyses Part 17

I created a better figure!

In this script, I first separated my peptide abundance data by site. Then, I averaged peptide abundance at each site. I saved the datafram as a .csv here.

Using the dotchart function in R, I created two plots that I can use as the cornerstone of my DNR paper. The first figure I made included all of my data, which Brent suggested.


Figure 1. Average peptide abundance data across all sites.

Each protein is coded with the same color, and with different shapes indicating the constituent peptides. I also used this color scheme to indicate protein function:

  • Heat shock: orange/red
  • Acid-base balance: purple
  • Drug resistance: pink
  • Fatty acid metabolism: yellow
  • Carbohydrate metabolism: brown
  • Cell and growth maintenance: green

I didn’t add a legend, since I thought I’d first get feedback on the figure, then make it better. The downside with this graph is that there are 37 peptides at 5 different locations in this one figure, so it’s a little busy.

I decided to narrow down the figure to the peptides were differentially expressed. I highlighted the differentially expressed peptides in the figure below.


Figure 2. List of differentially expressed peptides.

I noticed that heat shock protein wasn’t included in this list since I revised p-values! There are 16 total differentially expressed peptides. I then used this dataset to remake my figure.


Figure 3. Differentially expressed average peptide abundance across all sites.

This figure is cleaner. We’ll see what Steven, Brent and Emma think at our meeting on Tuesday!

// Please enable JavaScript to view the comments powered by Disqus.

from yaaminiv.github.io http://ift.tt/2BEdSKP

Grace’s Notebook: Tuesday, November 7, 2017

Yesterday, I went with Laura to Manchester to help clean and de-symbiont about 1600 oysters. There were 7 or 8 of us working, so we got it all done in a little over three hours! And, it was thankfully a beautiful, sunny (cold) day. I scraped off tons of barnacles, limpets, algae, and red worms.

Today, I worked more on the DIA analyses protocol. Specifically, Step 2. Some issues came up, so Steven and Sam have both been helping address them so that we can move forward. I am now looking at 3a and 3b. My task is to work on understanding what these steps mean, and then Steven said we’ll wait and see what we can do about running Pecan. Apparently, it’s difficult to use, and may have to be sent to Emma to run.

I then revisited this GitHub Issue and took some images to provide an example of what this project would be like if I followed suit with what I’ve done in the past.

Sam, Steven, and I (I mostly stood watching) worked on connecting the titrator to the computer (Swan). It was complicated, but I think they got it to work. My task for tomorrow is to make the NaOH (0.1 mol/L) solution in the titrant bottle so that we can hopefully start moving forward with the titrator!


Yaamini’s Notebook: Remaining Analyses Part 16

Tying up (more) loose ends

Brent and Emma suggested a few things to me many moons ago:

  • See if there’s a site-level bare vs. eelgrass effect
  • Correct my p-values for multiple comparisons

A few weeks later, I finally did it.

Bare vs. eelgrass effect

R script

I took my data and broke it up by site. I then conducted an ANOVA at each site testing differences in protein abundance between bare and eelgrass habitats. I wrote out the ANOVA F-statistic, original p-value, and Benjamini-Hochberg corrected value in the following data tables:

Case Inlet

Fidalgo Bay

Port Gamble Bay

Skokomish River Delta

Willapa Bay

I also created a bunch of boxplots, which can be found in this folder. I did not find any site-level effects! I’m honestly glad that there are no site-level effects. That would make my story waaaaay more complicated.

Correcting for multiple comparisons

R script

According to this handy dandy stats website, doing a bunch of statistical tests can lead to some significant results just by chance. That’s not good. With the Benjamini-Hochberg method, you can set a false discovery rate, then only count p-values that mee that FDR criteria. The FDR should be set ahead of time, before looking at the data. Since this is an exploratory study, I don’t want to set a stringent FDR and miss something. But I also don’t want to set my FDR so high that everything is significant. I settled on 10%.

In R, I used the function p.adjust to take the p-values I got from the ANOVA and modify them with the B-H method. Any p-values less than my FDR are considered significant. For these peptides, I can look at the Tukey HSD p-values, which are already adjusted.

Here’s my revised table. I have 13 peptides that are significant!

My next step is to incorporate all of the peptide abundance information into a succinct figure.

// Please enable JavaScript to view the comments powered by Disqus.

from yaaminiv.github.io http://ift.tt/2odhvij

Yaamini’s Notebook: Reproductive Output Analyses

i.e. My grand return to RStudio

It’s been a while since I did any stats-related! I tackled the preliminary analysis of the egg production and hatch rate data from Manchester. Since my original data was in an .xlsx file with several different tabs, I converted the relevant tabs into separate .csv files. I then created a new R script and started coding some ANOVAs.

Egg Production


There were significant differences in egg production between treatments (F = 25.87017, p = 0.00112206). Females exposed to heat shock produced more eggs than females exposed to either low or ambient pH treatments (HS-A = 0.0022743; L-HS = 0.0016528).

Hatch Rate




There was only a significant difference between treatments when taking female treatment into account (F = 5.606537, p = 0.01039109). This was due to a difference between low and ambient pH hatch rates, with low pH hatch rates being less than ambient pH hatch rates (L-A = 0.0111872). This is what I suspected when I first looked at this data!


Females exposure to low pH treatments could explain lower larval hatch rates #MaternalEffect Since there was no difference in the number of eggs produced between females exposed to low and ambient pH treatments, perhaps low pH eggs are lower quality. There could also be an epigenetic effect at play (cue MBD-seq preparation…or at least cue bringing this finding to Steven’s attention and hoping there’s money in the budget).

Once I sort out my histology data, I think I’ll have a story to tell for NSA.

// Please enable JavaScript to view the comments powered by Disqus.

from yaaminiv.github.io http://ift.tt/2o6Jdxb

Laura’s Notebook: (Mid-)February Goals


  • Revise DNR paper based on SR & GC feedback, particularly the Discussion + results
  • Create final plot for DNR paper
  • Create list of datasets, tables, etc. to include in supplementary content
  • Isolate DNA from gonad histology
    • Test out protocol on 3-5 samples, varying homogenization steps, through assay – 1 day (Feb. 13 or 14)
    • Run protocol on 18 samples – 1 day (Feb 23)
  • Shear DNA to 500bp (do I need to have Mackenzie help me? Probably. Schedule this with her.) 1/2 day (Feb 26)
  • Perform methylated DNA enrichment – 2 days (Feb 27 & 28)
  • Finish AA presentation, practice.
  • Teaching stuff.
  • Sample @ Manchester & Mud Bay, Feb. 27th (exposed 7pm -> 1am)

from LabNotebook http://ift.tt/2F1jC0X

Laura’s Notebook: An Observation, 2017 Oly Data

Looking at the Oly data from the 2017 Oly project, I see a couple patterns. Check out the below image, where I simply identify the treatment within each population that had the MOST and LEAST of the following:

  1. Gonad ripeness post-OA
  2. Larval production
  3. % larval survival

The coding as as follows:

  • 6L: 6degC, low pH
  • 6A: 6degC, ambient pH
  • 10L: 10degC, low pH
  • 10A: 10degC, ambient pH

My observations:

  1. The group that spawned the most number of larvae (normalized per oyster) consistently had the poorest larval survival across all populations.
  2. The South Sound F1 and F2 groups have identical patterns, indicating a genetic component in how the environment impacted phenotype.
  3. Some similar patterns between Hood Canal and Fidalgo Bay, but less so than the SS F1 and F2

most least

from LabNotebook http://ift.tt/2o0MFJO

Laura’s Notebook: Gonad histology stats, all treatments

Here are a bunch of contingency tables showing gonad stage and sex pre- and post-OA treatment. The p-values are from either Chi-squared test or Fisher’s exact test. They are not corrected for multiple comparisons.







from LabNotebook http://ift.tt/2nWhakT