Experimental Design:

• 4 populations of Fundulus grandis were collected across the Gulf of Mexico:

• Clean sites: Leeville, LA (LA-REF 29°15'24.6"N 90°12'51.3"W) and Gangs Bayou, Galveston, TX (TX-REF; 29°15'30.31"N; 95°54'45"W)

• Chemically Polluted sites: Grand Terre Island, LA, contaminated by the Deepwater Horizon oil spill (LA-DWH; 29° 16' 22.4472'' N 89° 56' 41.46'' W) and polluted Superfund site at Vince Bayou, Houston, TX (TX-SF; 29°43'10"N; 95°13'13"W)

• Adult fish were cultured in lab conditions for 1-2 generations before experiment

• Parent treatment: Adult fish were exposed to WAF or clean/control water for ~60 days before spawning

• Embryo exposure dosage: Embryos were exposed to WAF 00%, 10%, 32%, or 56% for 21 days

• Embryo developmental stage: Embryos were sampled for RNA-seq at stages 19, 28, 35, or hatch

  • Leftover embryos that were unhatched at the end of 21 days were collected as non-hatched (NH)

• RNA-seq data for:

  • ARS: Aquatic Research Station, aquacultured pop in LA
  • GT: Grand Terre, recently oiled pop in LA
  • VB: Vince Bayou, industrial pollution pop from Houston Ship Channel
  • GB: Gang's Bayou, clean ref pop near Houston Ship Channel

• Up to 5 replicates were sequenced for each treatment

• File naming convention:

  • [SampleNumber, 4 digits] ARS [ParentTreatment, 1 letter] [Embryo exposure %, 2 digits] [Developmental Stage, 2 digits] [ReplicateNumber 1-5]
  • Sample Number: unique number assigned to each embryo sequenced
  • ARS: population of F. grandis
  • Parent Treatment: Exposed (E) or Control (C)
  • Embryo exposure: 0, 10, 32, or 56%
  • Developmental Stage: 19, 28, 35, or HA
  • Replicate Number: Single digit, 1-5

• Samples 75, 145, 146, and 163 should be excluded from analysis due to barcoding error (see "Identifying samples that didn't get sequenced in the first lane trial")

General Pipeline:

• [Quality control with FastQC]
• [Trim short reads with Trimmomatic]
• [Map and quantify reads with Salmon]
• [Differential expression analysis in R]

Quality control with FastQC

FastQC is a raw read quality assessment command line tool. I used the following parameters to QC reads before and after trimming:

fastqc file --noextract -o ToDir		

Trim short reads with trimmomatic

Reads were trimmed lightly using Trimmomatic command line tool and NEBNext adaptor .fa file.

[0002_trimmomaticNEB.sh] link to the script

The .fa files are located at:

~/programs/Trimmomatic-0.36/adapters

Merge trimmed reads

Notes: looks like I copied the trimmed reads into intermediate mergedfq directories before merging and dumping into OUTDIR=~/niehs/Data/trimmed_data/mergedfq

Map and quantify reads with Salmon

Reference transcriptome:

• Lisa Cohen concatenated two [reference transcriptomes by Don Gilbert] (http://arthropods.eugenes.org/EvidentialGene/killifish/genes/kfish2rae5/) that contain alternate gene model information. I used that concatenation, found here:

    ~/niehs/refseq/kfish2rae5g.mrna.combined
  

• Trimmed and merged data located:

    ~/niehs/Data/trimmed_data/mergedfq/
  

Reads were ultimately mapped to Don Gilbert's reference transcriptome, fmd mode.

Differential Expression Analysis in R

Gene-level counts from transcript abundances were estimated using tximport in R.