Sam's Notebook

Daily Bits - September 2022

by Sam White 5 min read

20220930

  • Continued to review Circos documentation.

    • Thought about usefulness (uselessness?) of links for our use case. Possibly interesting to link genes with same GOslims?

  • Science Hour

  • Worked on CEABIGR/Circos organization. Should Circos be in it’s own branch? Does this simplify things or complicate things for usage? Better way to keep recommended Circos organization, but setup working directories outside of Circos program install location?

20220929

20220928

20220927

  • Worked extensively on Circos stuff for CEABIGR project. Had to fix some R code, as well as file formatting, bash scripts, and learning how to manipulate the circos.conf file. Actually generated a plot:

    Circos plot showing the C.virginica chromosome, NC_035781.1, as the black outer ring, with control female mean gene expression values (black inner ring) and exposed female mean gene expression values (green inner ring)

20220926

  • epidiverse/snp pipeline completed, so transferred data and wrote up notebook entry.

  • Generated CEABIGR mean gene expression files, formatted for Circos usage.

20220923

  • Resubmitted epidiverse/snp pipeline on Mox for Crassostrea virginica (Eastern oyster) because I forgot that the pipeline has an aritficial cap of 10 BAMs. Edited the nextflow.config file to increase cap to 50 and added code in the bio_DNA-methylation.md doc to count the number of BAMs and specify the epidiverse/snp command to use that number of BAMs. Prevents the aritficial limit from having any impact.

  • Science Hour. Helped Matt with some Mox/Trinity (system $PATH stuff).

20220922

  • Updated ceabigr predominant isoform R Markdown (GitHub) to generate files for control/exposed female samples.

  • Due to issue with R loading a package and getting this error: libgsl.so.23: cannot open shared object file: No such file or directory, I came across some suggestions that if your current version of R was built with a previous version of Ubuntu (which happens to be my case, since I upgraded Ubuntu earlier this month), I decided to upgrade and build R. Took awhile to compile… I also ended up having to re-install packages. Ugh!

  • Initiated epidiverse/snp pipeline on Mox for Crassostrea virginica (Eastern oyster) Bismark BAMs, per this GitHub Issue.

20220921

20220920

  • ceabigr meeting with Yaamini. Decided to generate list of genes with differing predominant isoforms between females and males - these were the only two comparisons Steven had produced so far. List will also use a binary system (i.e. 0 or 1 to indicate no different or different, respectively).

20220919

  • Did I finally fix the SBATCH script for geoduck Hisat2 alignments? It’s looking that way…

  • CEABIGR Circos stuff

    • awk 'BEGIN {OFS="\t"} {print "cvir"$1, $2, $3}' C_virginica-3.0_Gnomon_genes.bed > C_virginica-3.0_Gnomon_genes.bed.circos

    • Continued reading documentation.

  • Got distracted exploring how to get a list of all existing GO IDs. The idea being to then map all GOslims to the GO IDs and create a “flat” file that can be used for joining. Mostly reading about using the GO.db library in R and how to extract information.

20220916

20220915

20220914

  • Installed Circos on my computer (VM Ubuntu 22.04LTS):

    • Couldn’t install needed library:

      sudo apt-get -y install libgd2-xpm-dev
E: Unable to locate package libgd2-xpm-dev

      Possible fix is: sudo apt-get -y install libgd-dev

    • Missing Perl modules:

      sam@computer:~/programs/circos-0.69-9/bin$ ./circos -modules
ok       1.52 Carp
ok       0.45 Clone
missing            Config::General
ok       3.80 Cwd
ok      2.179 Data::Dumper
ok       2.58 Digest::MD5
ok       2.85 File::Basename
ok       3.80 File::Spec::Functions
ok     0.2311 File::Temp
ok       1.52 FindBin
missing            Font::TTF::Font
missing            GD
missing            GD::Polyline
ok       2.52 Getopt::Long
ok       1.46 IO::File
missing            List::MoreUtils
ok       1.55 List::Util
missing            Math::Bezier
ok   1.999818 Math::BigFloat
missing            Math::Round
missing            Math::VecStat
ok    1.03_01 Memoize
ok       1.97 POSIX
missing            Params::Validate
ok       2.01 Pod::Usage
missing            Readonly
missing            Regexp::Common
missing            SVG
missing            Set::IntSpan
missing            Statistics::Basic
ok       3.23 Storable
ok       1.23 Sys::Hostname
ok       2.04 Text::Balanced
missing            Text::Format
ok     1.9767 Time::HiRes

      Needed to install cpanm:

      apt-get install cpanminus

      Then:

      sudo cpanm Clone Config::General Font::TTF::Font GD GD::Polyline List::MoreUtils Math::Bezier Math::Round Math::VecStat Params::Validate Readonly Regexp::Common SVG Set::IntSpan Statistics::Basic Text::Format

      Got me to this:

      sam@computer:~/programs/circos-0.69-9/bin$ ./circos -modules
ok       1.52 Carp
ok       0.45 Clone
ok       2.65 Config::General
ok       3.80 Cwd
ok      2.179 Data::Dumper
ok       2.58 Digest::MD5
ok       2.85 File::Basename
ok       3.80 File::Spec::Functions
ok     0.2311 File::Temp
ok       1.52 FindBin
ok       0.39 Font::TTF::Font
ok       2.76 GD
ok        0.2 GD::Polyline
ok       2.52 Getopt::Long
ok       1.46 IO::File
ok      0.430 List::MoreUtils
ok       1.55 List::Util
ok       0.01 Math::Bezier
ok   1.999818 Math::BigFloat
ok       0.07 Math::Round
ok       0.08 Math::VecStat
ok    1.03_01 Memoize
ok       1.97 POSIX
ok       1.30 Params::Validate
ok       2.01 Pod::Usage
ok       2.05 Readonly
ok 2017060201 Regexp::Common
ok       2.87 SVG
ok       1.19 Set::IntSpan
ok     1.6611 Statistics::Basic
ok       3.23 Storable
ok       1.23 Sys::Hostname
ok       2.04 Text::Balanced
ok       0.62 Text::Format
ok     1.9767 Time::HiRes

      Successfully generated test images!

  • Tidied up a bunch of things in the SBATCH script for geoduck RNAseq HiSat alignments in order to get it to run properly. Currently waiting in queue…

  • Updated (circos_pgen_karyotype.sh)[https://github.com/RobertsLab/sams-notebook/blob/master/bash_scripts/circos_pgen_karyotype.sh] bash script to allow passing arguments for species abbreviation and FastA index file.

20220913

20220912

20220909

  • Science Hour and Pub-a-thon

  • Started working on SBATCH script for geoduck RNAseq HiSat alignments for eventual lncRNA id.

20220908

20220907

  • Computer maintenance: upgraded my laptop Ubuntu virtual machine from 20.04LTS to 22.04LTS.

  • ProCard reconcilation.

  • transferred geoduck RNAseq files to Mox in preparation for long non-coding RNA identification.

20220906

20220901

  • Continued to mess around with Mixomics R package tutorials. Here’s R Markdown used, as well as some of the plots that were produced (PCA and sparse PCA). Overall, sex is driving factor of differences in gene expression (FPKM) and average gene methylation.

Put together a quick R Markdown doc to go through the tutorial using our gene FPKM expression data and average gene methylation data (generated by Steven):

https://rpubs.com/kubu4/cvir-gonad-oa-mixomics_testing

Some screenshots of plots generated:

PCA plot comparing impacts of average gene methylation, OA exposure, and sex. Orange shapes are males. Blue shapes are female. Circles are control water conditions. Triangles are exposed to OA water conditions. Oranges (males) are all grouped together on the left side of the PCA plot. Females are all grouped together on the right side of the PCA plot.

Sparse PCA plot (sPCA) comparing impacts of average gene methylation, OA exposure, and sex. Orange shapes are males. Blue shapes are female. Circles are control water conditions. Triangles are exposed to OA water conditions. Oranges (males) are all grouped together on the upper left side of the sPCA plot. Females are all grouped together on the upper right side of the sPCA plot. A single, blue triangle is on the bottom right side of the sPCA plot.

Screenshot showing list of genes and the sPCA loading values contributing the most to the sPCA Comp 1 from above. List is followed by bar plot to visualize the loadings in the list above.

A Sparse Projection to Latent Structure (sPLS) plot of C.virginica gonad gene expression (FPKM) values and gene average methylation values. Blue letters are control water conditions. Orange letters are exposed OA water conditions. The letter 'F' represents females. The letter 'M' represents males. All females are tightly clusterd in the lower left corner. All males are tightly clustered in the lower right corner.