Most highly abundant transcripts (contigs with >1,000 reads).

*<p>Blastx were based on nr protein database; all accessions were from <i>Gossypium</i> species as ranked among 5 best hit scores except ADB54351 from <i>Allium.</i></p>**<p>the blastx annotation as FbLate-2 in <i>G. barbadense</i> ranked 11^t<i>h</i> (others were hypothetical protein), had 100% identity but only 25% coverage.</p><p>Number of reads of genotypes and fiber development stages are summed over 2 libraries each (<i>Gh</i> = Gh10+Gh22, <i>Gb</i> = Gb10+Gb22, 10 dpa = Gh10+Gb10, and 22 dpa = Gh22+Gb22).