Has anyone observed that when making paired-end libraries the yield of final PCR product is much less as compared to making single-read libraries? I...