Grace’s Notebook: Wednesday, September 20, 2017

I was only in today for a couple hours.

I went through this google drive folder (here) to see if any of the images match those in the O. lurida survey paper. They do, but only Figures 2, 3, 4, and 11-13. Figures 2-4 are missing the specifically labeled temperatures that are in the figures in the paper, and the site names are incorrect: they should be Southern, Northern, Hood Canal, and Central. I’m not sure what the resolution is on these images, but we’re trying to get figures that are 600 dpi.

Figures 1, 5-10, and 14 are still missing.

I then tried to work on the problem that occurred at the end of Monday with my blast to GOslim notebook: here. I really am not sure what’s going on, and Sam is trying to help me… but I don’t quite understand. It appears that the joining is working, but visualizing it isn’t working …? I don’t know.

Screen Shot 2017-09-20 at 3.24.52 PM.png


Grace’s Notebook: Monday, September 18, 2017

The Roberts Lab logo stickers arrived today! IMG_4324.JPG


Today, Steven and I checked on the BLAST of cgigas and swissprot that we started on Friday afternoon. It completed the process on Saturday 1:31 AM (took about 7 hrs).

Screen Shot 2017-09-18 at 1.56.48 PM

Next, we used this notebook template: here – to annotate the BLAST output with GOslim terms. I made a copy of the notebook and changed the working directory to my folder on the desktop of Genefish.

Link to my practice notebook: here. Things went well until the last step. Something isn’t allowing for the files to be joined. Created a GitHub issue for input.


Screen Shot 2017-09-18 at 3.01.15 PM

I also started a proteomic data inventory google spreadsheet. Link: Here.

Grace’s Notebook: Friday, September 15, 2017

I spent time moving image folders to Owl. Still not completely updated and still have some “readme” files to write and add. Link to folder: here.

I also set up a BLAST with Steven on the gene fish computer. We created a database using uniprot, and then blasted it against C. gigas. Link to make-database notebook: here. Downloaded the uniprot-swissprot from: here. swissprot is annotated uniprot.

Link to blast notebook: here. It will be let to run through the weekend and we’ll check back on it on Monday.

Grace’s Notebook: Wednesday, September 13, 2017

  • Today I submitted the revisions we made to the “Evidence of Ostrea lurida Carpenter, 1864, population structure in Puget Sound, WA, USA” paper.
  • Ordered 2″x2″ clear, round, lab logo stickers! 🙂
  • Set up the new peristaltic pump. Will work on how to attach filters later.
  • Working on figuring out what these data files – link: here –  in Rhonda’s folder are.

Will continue to move image folders to Owl, as well as work on compiling google sheets, as requested.

Yaamini’s Notebook: SRM Analysis Part 6

Retracing my steps

Since I don’t know where things are going wrong, I’m going back to everything I’ve done and making sure things are okay. The first thing I did was go through my R script to see if there was anything that was incorrectly assigning samples to data or eliminating them. I couldn’t find anything, so I started writing a new R script for an NMDS plot after averaging technical replciates. As I was writing this script, I found that my dataframe technicalReplicates were having sample names alternating between -1 and -2, instead of just -1. I went all the way back and found that my exported report from Skyline did not have samples O106-1 and O71-1!

I added in the RAW files for my missing replciates, files 10 and 28. These two samples didn’t have any missing data, so I’m not sure why they weren’t included in the first place! I also decied to readd the files that originally displayed no chromatograms just to see if there might have been an importing error: files 3, 16, 22 and 34. Files 3, 22 and 34 displayed data so I kept them, but 16 did not. I removed it and tried imorting it a third time. It still didn’t work, so I removed the file.

While I was in Skyline, I decided to check my PRTC peptides as well. I started by seeing what peaks were present in each PRTC peptide chromatogram. They all had one very clearly defined peak. The only peptide that did not have any peak picked was TASEFDSAIAQDK.










*Figures 1-9. Peaks present in PRTC peptides.

Using this document to predict retention times, I ensured that all of the proper peaks were picked.


Figure 10. Proper peak selected for PRTC peptide.

I then (accidentally) saved my revised Skyline file under the same name and overwrote the original file in OWL. The Skyline document can be found here.


*Figure 11. Export settings for revised data.

I made sure the sequence file contained information for all of my samples, with the exception of O12 (one technical replicate is file 16, which has no data. I had to manually add the protein name CHOYP_ACAA2.1.1 m.30666 to the file since that protein name was the only one that wasn’t exporting. Finally, I made sure I was missing no data when going through my R scripts. I resaved dataframes and images that were affected by the readdition of my samples. My NMDS clustering and ANOSIMs are still not producing anything desirable. I’m going to focus on making protein bar graphs for PCSGA instead.



Grace’s Notebook: Tuesday, September 12, 2017

Today I started out by taking images of Yaamini’s C. gigas post-OA exposure histology. Link to the image folder: here.

Then, I read through the 2015 Oyster Larvae Proteomics paper (link: here) and am working on determining the day timeline, as well as the methods in relation to what was done with 4 proteomics samples.

I am also beginning the process of backing up all the images I’ve taken of oyster and geoduck seed, and histology slides to OWL. Link to my folder: here.

Tomorrow, I will continue with the paper, transfers to OWL, and I will test out how to use our new peristaltic pump with water, so that when/if the time comes to use it, we’ll already know.

Yaamini’s Notebook: SRM Analysis Part 5

Everything sucks.

R script

I normalized my data using TIC values, saved in my sequence file. After dividing all of my area values by TIC content, I remade my NMDS plot. Things did not look much better :disappointed:


During my meeting with Julian, he suggested I calculate the distances between ordinations for each technical replicate so I have a quantitative way of pulling out bad technical replicate clustering. For my bad NMDS, the distances didn’t look great :cold_sweat:


Emma said she’s never seen such bad technical replication. Will I have yet another failed experiment? Will I be able to present anything at PCSGA? Who could never be sure :sob: