Fig 1.
Schematic representation of the Computel algorithm for mean telomere length estimation.
Computel takes whole-genome NGS short-reads as input; maps them to the telomeric index built based on user-defined telomeric repeat pattern and the read length; and calculates the mean telomere length based on the ratio of telomeric and reference genome coverage, the number of chromosomes, and the read length.
Fig 2.
An example of sequence structure of telomeric index for telomeric read alignment.
Telomeric pattern is “TTAGGG” (human); read length = 20 nt; seed length (min.seed option) = 10 nt. The top read contains a non-telomeric region, which will be aligned to the non-telomeric tail of the index, the rest of the reads are six possible cyclic permutations of pure telomeric repeats.
Fig 3.
Correlation between actual and estimated mean telomere lengths.
S—single-end reads, P—paired-end reads. Estimation of mean telomere length was performed with reads generated from 200 kb length region of human chromosome 1, with telomeres attached to both its ends with lengths sampled from a normal distribution with mean 10 kb and SD 7 kb. The minimum-maximum range of the generated telomere lengths were: 194.5–21138 bp for single-end reads, and 387.5–24169.5 bp for paired-end reads. The read length, insert size and fold coverage ranges are described in the Materials and Methods.
Table 1.
Comparison of performance of Computel and TelSeq in mean telomere length estimation from synthetic data.
Table 2.
Log ratios of telomere length estimates by Computel and TelSeq compared to qPCR for five neuroblastoma (D) and matched normal tissue (N) samples.
Fig 4.
Mean telomere length estimates for osteosarcoma and matched normal tissues by qPCR, Computel and TelSeq.
SJOS002_D, SJOS004_D—osteosarcoma tissue samples; SJOS002_N, SJOS004_N—paired healthy tissue samples.