3.10 Transcriptome analysis after T. laevis and T. controversa infection
In order to known the transcriptional changes in wheat anther developmental stages after T. laevis and T. controversainfection, we analyzed anthers of different lengths including; 0~150 μm, 150~400 μm, 400~700 μm, 700~1100 μm and 1100~1300 μm by RNA-Seq. The raw reads were obtained by removing the adaptor sequence, ambiguous bases, and low-quality sequences. For T. laevis , approximately, 70.74 (control-2-3) to 91.88 (control-2-2) million clean reads were obtained from control samples and their mapping rate were 85.76% to 69.10%, respectively, and clean reads in T. laevis infected samples were 75 (infected-5-2) to 108.29 (infected-2-3) million reads and their mapping rate were 68.57% to 83.39%, respectively. The Q20, Q30, GC content and mapped reads of above mentioned transcripts ranged from 98.51 to 98.76 %, 95.56 to 96.17%, 52.89 to 54.98% and 51.94 to 90.30 million, respectively (Table 1 ). For T. controversa , approximately 89.77 (control-2-3) to 104.68 (control-2-2) million clean reads were obtained from control samples and their mapping rate were 87.26 % to 87.04%, respectively, and clean reads in T. controversa infected samples were 88.30 (infected-1-1) to 121.96 (infected-5-3) million reads and their mapping rate were 82.62 % to 83.71%, respectively. The Q20, Q30, GC content and mapped reads of above mentioned transcripts ranged from 98.24 to 98.52%, 94.78 to 95.46%, 53.16 to 54.77 % and 74.33 to 92.40 million, respectively (Table S2 ). In order to clarify the reproducibility between samples, we analyzed and calculated the individual correlation coefficients between samples through principal component analysis (PCA). The biological replicates of each treatment were clustered together to prove that there was good repetition between replicates and there were differences between treatments for T. controversa (FigureS3A ) and for T. laevis (Figure S3B ).