RSEM and MultiDEA read count estimations.

Read count estimation computed by RSEM and MultiDEA for each isoform in HELLS genomic location, for each sample and for each dataset. Despite the diffuse overlapping among the isoforms, the Bayesian model included in RSEM tends to prefer only a small subset of isoforms, changing in each dataset. MultiDEA clearly shows that the only isoforms to achieve a significant number of uniquely mapping reads (>20) are XR_007061960 and ENST00000419900, while all the other reads map to more than one reference transcript. RSEM and MultiDEA columns list the isoform expression estimation computed by each tool; [A,B,C,D] is the multiread detail computed by MultiDEA (trapezoid vertices, A = uniquely mapping reads, B = reads having the isoform as unique best match; C = reads having that isoform as non-unique best match; D is the total number of reads that map to that isoform). Isoforms are sorted by start position on chr10; XR_007061960 is an additional isoform of HELLS proposed by RefSeq and ENST00000432120 is a lncRNA that overlaps HELLS. (XLSX)