This dataset was generated using RNAseq profiling of matching normal, tumor and metastasis tissues from eight colorectal cancer patients.
Surprisingly, the difference is relatively small, such that non-matched samples, whether between normal or tumors, are still considerably more different than matched normal-tumor.
Through matched normal-tumor pairs, it is now possible to identify genome-wide germline and somatic mutations.
Using IBS on two datasets of matching gastric cancer samples (SNP6 and NGS), we show that clustering of matched normal-tumor samples can be used to assess the 'matchedness' of pairs.
While the approach in NGS has been rigorous in the generation of data as well as analysis, we note that there is a lack of quality check to assess if the matched normal-tumor pair was indeed matched.
As IBS has been traditionally used in genetics studies to identify variants in populations, it is thus novel to apply the idea in cancer samples, especially in this particular context of assessing matched normal-tumor pairs.
In some of the studies, matched normal-tumor pairs were utilized to identify germline and rare somatic mutations [4], [6], [7], while in other studies, matched tumor-tumor (e.g. primary versus metastatic) were used [1].
All variants from the matched normal-tumor pairs that were unique in the tumor sample were called as Somatic (using Samtools mpileup funtion).
As can be seen in Table 3, the percentage of incompatible genotypes in the 88 passed, matching normal-tumor pairs (after excluding the pairs mentioned above) was low in all the collections, but tended to be higher in FFPE pairs than fresh frozen pairs (average 0.1% versus 0.01%).
In addition to paired normal and tumor tissue, matched normal near tumor and adenoma tissue samples were examined.
