SupplementaryMaterials and Methods
Next-generation deep-sequencing target accession numbers (NGS)
PCR primer design for next-generation deep-sequencing
PCR amplification protocols for next-generation deep-sequencing
PCR amplification 96-well plate layout
PTP layout and emPCR condition
Data on bead enrichment and PTP bead loading
Coverage considerations
Next-generation deep-sequencingtarget accession numbers (NGS)
The primer design for sequencing of genomic DNA was based on the Ensembl accession numbers given in Supplementary Table 1 (Ensembl 53: Mar 2009).
PCR primer design for next-generation deep-sequencing
Supplementary Table 2 lists corresponding primer pairs for 31 amplicons representing TET2, CBL, and KRAS.The specific fusion 454 adaptor sequences were:
- primer Afusion sequence: 5' CGTATCGCCTCCCTCGCGCCATCA G 3'
- primer Bfusion sequence: 5' CTA TGC GCC TTG CCA GCC CGC TCA G 3'
In total, three distinct molecular barcode sequences were incorporated into the primer sequences after the TCAG sequencing key and before the gene-specific sequence:
- MID-1: ACGAGTGCGT
- MID-2: ACGCTCGACA
- MID-4: AGCACTGTAG
PCR amplification protocols for next-generation deep-sequencing
PCR protocols were performed with ~40ng template genomic DNA. All PCR master mixes were prepared according to the manufacturer’s recommendations using the FastStart High Fidelity PCR Kit (Roche Applied Science, Mannheim, Germany). PCR reactions were performed usinga 96-Well GeneAmp® PCR System 2720 or 9700 instrument (Applied Biosystems, Foster City, CA) including a final cooling step at 4°C after final elongation (Supplementary Table 3).
PCR amplification 96-well plate layout
One preconfigured primer plate allowed single-plex parallel PCR reactions of three patients including three different MIDs. After purification and quantification, the resulting amplicon products were pooled in an equimolar manner. These amplicon pools were subsequently diluted to a concentration of 1 x 109 molecules/µl. For the subsequent emPCR assay three patients with different MIDs were combined into one lane pool (Supplementary Figure 1).
PTP layout and emPCR condition
Prior to emPCR, single lane pools were diluted stepwise to a final concentration of 4 x 106 molecules/µl. For each emPCR setup, 1.05 µl of this library was used which corresponded to 2.1 copies per bead. For all 8 lanes of the PTP, A (forward) and B (reverse) beads were processed separately. After breaking and enrichment A and B beads were combined for subsequent sequencing (Supplementary Figure 2).
Data on bead enrichment and PTP bead loading
As given in Supplementary Table 4, bead enrichment rate and PTP loading schemes differed between the respective centers. According to the study protocol, ideally 8% - 15% of beads would have been expected for enrichment. Subsequently, in total 340,000 beads, e.g. 170,000 A beads and 170,000 B beads were instructed as recommendation to be loaded per PTP lane. It is noteworthy that bead loads varied from 71,250 to 400,000 per lane.
Coverage considerations
In Supplementary Table 5, the variant CBL C401Y (exon 8; 268G>A) as detected in sample #16 is highlighted. For each center, the read direction is given along with the mutational burden as calculated by the onboard 454 GS Amplicon Variant Analyzer software version 2.3. Forward and reverse read coverage values and total reads are given. Interestingly, this mutation with <2% frequency was only detected in 6/10 centers. Centers 1, 6, 9 and 10, which detected the mutation consistently, achieved a coverage of over 250-fold in each sequencing direction. In centers 2, 3, 7, and 8, only reverse reads detected the mutation. In these four centers, the respective coverage for the forward reads ranged from 153 to 368. Thus, although centers 3 and 7 had achieved a coverage > 250-fold, the mutation was not detected. In contrast, centers 4 and 5 did detect the mutation even though in at least one sequencing direction the coverage was less than 250 reads.
1