Analysis | CNV insertion could be detected in Long Insert sized Paired End (LIPE) data on the basis of compressed pairs which indicated insertion of certain sequence between pair reads. The average insert size of the LIPE library was 2,700 bp and the standard deviation (SD) was about 300 bp. Based on these figures, a compressed pair was defined as one with insert size shorter than 2,000 bp (< mean - 2 SD). Similar to method used to detect deletion, compressed pairs were defined to be clustered if one compressed pair was within 3,300 bp (mean + 2 SD) to neighboring compressed pair. To remove redundant pairs present in LIPE library, two or more pairs with both reads aligned to exactly the same position of reference genome were defined as one pair and named de-redundant. If 10 or more deredundant compressed pairs were found in one cluster, we defined this region as a candidate region for amplification. |