Figure 2

Experimental and computational workflow. (A) (1) Harvestingembryos after fertilization and subsequent aging all took place at18°C. (2) After lysis and homogenization of embryos usingInvitrogen lysis buffer and a pestle, lysates were stored at −80degrees. (3) Protein-coding mRNA species were selected directly usingthe mRNA DIRECT kit from Invitrogen. (4) Sequencing libraries wereprepared using the ScriptSeq kit from Epicentre/Illumina. The kit hasfour major steps: RNA fragmentation, cDNA synthesis, 3’-terminaltagging, and PCR amplification. Separate index tags were added to eachtimepoint during PCR amplification. (5) Libraries were size selectionusing a 2% pippin prep from Sage Science to 450 bp. (6)Quantification of libraries with Bioanalyzer and qPCR. (7) All sampleswere pooled into a single lane. (8) Sequencing was performed on anIllumina HiSeq 1000 with version 3 chemistry. (B) (1) Wet-labexperimental methods to isolate, prepare, and sequence the RNA result inraw reads. (2) Quality control removes adapter sequences, artifacts,ribosomal and mitochondrial contaminants; trims the GC-content bias; andadaptively trims the low quality sequence bases. (3) De novotranscriptome assembly using Trinity or Velvet-Oases. (4) Thetranscriptome assemblies are compared using a variety of metrics. (5)Reads are normalized prior to quantification and other downstreamanalyses. Reads for specific transcripts are normalized based onfragments per kilobase of exon per million fragments mapped (FPKM) andspike-in measurements.