Fig. 1

Computational framework. Transcriptome-wide analysis tools comprised large-scale conserved domain search using Batch CD-Search, RNA secondary structure prediction using RNAfold, functional annotation by Mercator, and coding potential assessment using CodeWise. These tools, in conjunction with co-expression network analysis by ArrayMining and signal transduction pathway analysis in Beacon, were used for detailed data analyses of a set of 2,938 transcripts. This population afforded the opportunity to identify candidate SVs transcribed from the same gene with potentially distinct functions in different stages of soybean embryo development. Each of the 1,395 genes had more than one transcript significantly differentially expressed during time-course of soybean embryo development (FDR < 0.05), leading to the detection of 2,938 transcripts. Tools are shown in black, inputs in blue, and outputs in green