Expression levels of the individual TE superfamilies were calculated by averaging the TPM values among replicates of each sex and then summing the average TPM of all contigs annotated to each superfamily. For TE superfamilies detected in both the genomic and transcriptomic datasets, we tested for a relationship between genomic abundance and expression levels in the gonads of each sex using linear regression on log-transformed data.
To identify differentially expressed contigs between testes and ovaries, DESeq2 (Love et al., 2014 (link)) was used with an adjusted p-value cut off of 0.05. Among the 15,011 total differentially expressed transcripts between testes and ovaries (including TEs, endogenous genes, and unannotated contigs), 869 were TEs, representing 18 superfamilies and other unknown TEs. Superfamilies with fewer than 10 differentially expressed transcripts between testes and ovaries were removed, leaving nine superfamilies; for each, we tested for a difference in expression between testes and ovaries using a t-test.
To identify differentially expressed contigs between testes and ovaries, DESeq2 (Love et al., 2014 (link)) was used with an adjusted p-value cut off of 0.05. Among the 15,011 total differentially expressed transcripts between testes and ovaries (including TEs, endogenous genes, and unannotated contigs), 869 were TEs, representing 18 superfamilies and other unknown TEs. Superfamilies with fewer than 10 differentially expressed transcripts between testes and ovaries were removed, leaving nine superfamilies; for each, we tested for a difference in expression between testes and ovaries using a t-test.
Free full text: Click here
