Work in Progress

Whitehead, A. (2010). The evolutionary radiation of diverse osmotolerant physiologies in killifish (Fundulus sp.). Evolution, 64(7): 2070-2085. https://whiteheadresearch.files.wordpress.com/2012/05/whitehead-2010-evolution.pdf

• Goal: What are the gene expression differences between 3 treatments, by species?

• 3 Genera: Fundulus, Adinia, Lucania

• 12 species

• Sequencing done on 3 flowcells, 3 sample replicates per treatment

• 3 treatments/physiologies: freshwater (FW), brackish water (BW), transfer (brackish->fresh, sacrificed after 1 day)

• Data stored on farm.cse.ucdavis.edu here:

  /home/nreid/rnaseq/rawdata/

in three separate flowcells:

141203_HS4B, 
flowcell C5NU8ACXX

141212_HS2A,
flowcell C5VPYACXX

141126_HS3A,
flowcell C5NGRACXX

Basecall stats here:

/home/nreid/rnaseq/rawdata/141212_HS2A/Basecall_Stats_C5VPYACXX.zip
/home/nreid/rnaseq/rawdata/141203_HS4B/Basecall_Stats_C5NU8ACXX.tar
/home/nreid/rnaseq/rawdata/141126_HS3A/??

Want to make a summary chart for each genus, species, treatment, sample:

* flow cell
* lane
* read number
* file numbers

Need to combine files from same lane (because default demultiplexing setting was used, threshold max # reads splitting into seperate files)

• info on using slurm: http://wiki.cse.ucdavis.edu/support:hpc:software:slurm

• files named, e.g.

  AWJRDD003_L_parva_transfer_3_TCTCGCGC-TATAGCCT_L006_R1_002.fastq.gz AWJRDD002_L_goodei_transfer_3_CGCTCATT-CCTATCCT_L004_R1_002.fastq.gz AWJRDD001_F_similis_BW_1_TCCGCGAA-ATAGAGGC_L002_R1_001.fastq.gz

  flowcell#_Genus_species_treatment_sample#_barcode_lane_read_file#.fastq.gz

• 2 populations of heteroclitus: MDPL, MDPL

  AWJRDD002_F_heteroclitus_MDPL_transfer_2_ATTCAGAA-ATAGAGGC_L004_R1_002.fastq.gz AWJRDD002_F_heteroclitus_MDPP_BW_2_ATTCAGAA-CCTATCCT_L003_R2_001.fastq.gz

  flowcell#_Genus_species_population_treatment_sample#_barcode_lane_read_file#.fastq.gz

• Subgoal: Since there is an annotated reference for Fundulus heteroclitus, it is not clear with a multispecies dataset like this whether a reference-guided or de novo transcriptome assembly approach would be better for analyzing differential expression of transcripts for this experiment. We will look at both and compare.

Adapting the Eel pond khmer protocols by C. Titus Brown et al:

https://khmer-protocols.readthedocs.org/en/ctb/mrnaseq/

1. trim: https://khmer-protocols.readthedocs.org/en/ctb/mrnaseq/1-quality.html

   • Trimmomatic
   • combine lanes

2. diginorm: https://khmer-protocols.readthedocs.org/en/ctb/mrnaseq/2-diginorm.html https://github.com/dib-lab/khmer-protocols/blob/jem-streaming/mrnaseq/1-quality.rst

3. Trinity: https://khmer-protocols.readthedocs.org/en/ctb/mrnaseq/3-big-assembly.html

   • by species
   • memory allocation

4. annotation/evaluation with transrate: http://hibberdlab.com/transrate/

   • use annotated translated aa from reference transcriptome?

5. differential expression

   • salmon
   • edgeR, DESEq2, limma

6. comparative transcriptomics

from Dr. Noah Reid, has already run these:

• Fundulus heteroclitus reference: http://arthropods.eugenes.org/EvidentialGene/killifish/
• run .sh script in Noah's directory to expand and correct reference files: /home/nreid/popgen/kfish3/15.09.22.modify_gff.sh
• Send Don Gilbert email if questions
• gff reference version 20130322 we use slightly diffrent than public ncbi version
• avoid using ncbi set, a lot of effort already put in to use version 2rae5g before format changed for submitting to ncbi

1: run trimmomatic

15.02.17.trimmomatic_array.sh 

2: run bwa mem, merge and index bams

bwa_mem_paired_array.sh
bwa_mem_unpaired_array.sh
mergebams.sh
indexbam.sh

3: run featurecounts

15.11.13.featurecounts.sh

4: analyze counts in edgeR

osmotic_edgeR_script.R

5: make some plots

osmotic_edgeR_plots.R