Dear all,
I hope this email finds you well.
I prepared cfDNA libraries using ThruPlex TagSeq (Rubicon) starting with 25 ng cfDNA measured using Qbit. In parallel I also prepared similar liberties using matched normal gDNA sheared to 150-200 bp. I captured these libraries using an Agilent custom panel of 0.47 Mb following Rubicon’s instructions (5 cycles PCR). For post-capture I used 16 cycles PCR (as suggested for panels 1 kb-0.5 Mb). Libraries look good after capture.
To be able to run downsampling experiments to determine the amount of sequencing required to obtain a given duplication rate, or to obtain a given family size I sequenced cfDNA samples so that I obtained 110-125 million PE 2x75 reads (220-250 million reads) and 20 million PE 2x75 reads (40 million reads) using NextSeq500 high output.
(Note: I analysed the samples using Curio software (specialized to analyze data from TagSeq), by aligning with BWA-MEM and selecting for UMI containing reads.)
When I analysed the coverage I was surprised to see that while the off-target (as judged by the high number of reads on untargeted regions) was much higher in the cfDNA sample than in the normal sample, the duplication rate (which i expected to be higher) was not incremented whatsoever and, moreover, the family size was actually lower than in normal. I fund this extremely odd. The coverage analysis indicated that ~2.5 million families were formed in both samples.
My question, how can it be that the off target soared without increasing the duplication rate nor the family size?
Any help will be appreciated.
Luciano
I hope this email finds you well.
I prepared cfDNA libraries using ThruPlex TagSeq (Rubicon) starting with 25 ng cfDNA measured using Qbit. In parallel I also prepared similar liberties using matched normal gDNA sheared to 150-200 bp. I captured these libraries using an Agilent custom panel of 0.47 Mb following Rubicon’s instructions (5 cycles PCR). For post-capture I used 16 cycles PCR (as suggested for panels 1 kb-0.5 Mb). Libraries look good after capture.
To be able to run downsampling experiments to determine the amount of sequencing required to obtain a given duplication rate, or to obtain a given family size I sequenced cfDNA samples so that I obtained 110-125 million PE 2x75 reads (220-250 million reads) and 20 million PE 2x75 reads (40 million reads) using NextSeq500 high output.
(Note: I analysed the samples using Curio software (specialized to analyze data from TagSeq), by aligning with BWA-MEM and selecting for UMI containing reads.)
When I analysed the coverage I was surprised to see that while the off-target (as judged by the high number of reads on untargeted regions) was much higher in the cfDNA sample than in the normal sample, the duplication rate (which i expected to be higher) was not incremented whatsoever and, moreover, the family size was actually lower than in normal. I fund this extremely odd. The coverage analysis indicated that ~2.5 million families were formed in both samples.
My question, how can it be that the off target soared without increasing the duplication rate nor the family size?
Any help will be appreciated.
Luciano
Comment