W023
Mapping Parent-of-Origin Effects in the Chicken Transcriptome Using de-novo Transcriptome Assembly or Reference Genomes

Date: Tuesday, January 12, 2016
Time: 3:10 PM
Room: Royal Palm Salon 5-6
Fernando Lopes Pinto , Swedish University of Agricultural Sciences, Uppsala, Sweden
Mohammad Hossein Banabazi , Animal Science Research Institute of IRAN (ASRI), Karaj, Iran
Anna-Maja Molin , Swedish University of Agricultural Sciences, Uppsala, Sweden
Elizabeth R. Gilbert , Dept. of Animal and Poultry Sciences, Blacksburg, VA
Christa Honaker , Dept. of Animal and Poultry Sciences, Blacksburg, VA
Paul Siegel , Dept. of Animal and Poultry Sciences, Blacksburg, VA
Göran Andersson , Swedish University of Agricultural Sciences, Uppsala, Sweden
Leif Andersson , Uppsala University, Uppsala, Sweden
Dirk-Jan de Koning , Swedish University of Agricultural Sciences, Uppsala, Sweden
The Virginia Tech high-growth and low-growth selection lines of White Plymouth Rock chickens have been developed by more than 50 generations of divergent selection for body weight at 56 days of age. These lines have been extensively used for QTL mapping in order to reveal the genetic basis for the huge selection response. To further clarify the mechanisms for how QTLs control phenotypic differences, we measured allelic RNA expression ratios to detect differential expression between alleles derived from the two lines. Using whole transcriptome and genome sequencing, we set out to detect allelic imbalance in the RNA expression of 6 F1 progeny from reciprocal crosses between generation 54 parents from the high and low parental lines. Using RNA samples extracted from liver, hypothalamus and longissimus muscle, we generated circa 250 M (100 bp) RNA sequencing reads per F1 individual as well as a 25-fold coverage DNA sequence of each of the parents. This experimental design allowed us to detect parent-of-origin specific differential expression, since we could predict and test obligatory heterozygous loci in the F1 birds. Using the chicken reference genome sequence and transcriptome we could only map 30% of the RNA reads. We also had an excess (>70%) of allelic imbalance. Using de-novo transcriptome assembly we could map > than 70% of the RNA reads and we observed a clear reduction in the allelic imbalance.  For about 5400 candidate SNPs we detected statistical evidence for a parent-of-origin effect at  P < 0.001 for about 500 SNPs.  The annotation of the de-novo transcriptomes is the next challenge.