< Back

We use the GenDx NGSengine application for HLA typing and HLA report generation.

The HLA loci DRB3, DRB4 and DRB5 are paralogs of locus DRB1.  Thus, they share a significant fraction of their genetic sequence with DRB1. The NGSengine HLA Typing algorithm makes statistical estimates of the best matching genotype between sample data and the IMGT/HLA database. If the algorithm is unable to determine whether sequencing reads should be assigned to DRB1, DRB3, DRB4 or DRB5 it will display the message ‘Insufficient data’.

When NGSengine cannot determine if a specific read should be assigned to either DRB1 or DRB3, due to the high sequence similarity of the two loci, the read will not be mapped to either locus.  If a large number of reads cannot be unambiguously assigned to either locus, then NGSengine will display the message “Insufficient data” for locus DRB3.


Fig. 1

As shown in Fig. 1, the HLA typing algorithm correctly typed DRB1 as “DRB1*01:01:01:01, DRB1*03:01:01:01” for this particular sample.  There were 44,275 reads aligned to the IMGT/HLA database reference sequence and 46,503 total reads assigned to DRB1, for a mappability score of (44,275/46,503)*100 = 95%.  The HLA typing algorithm also aligned 18,028 reads to DRB3, from a total of 18,887 reads, for a mappability score of 95%.  However, the typing algorithm displays DRB3 with “Insufficient data”.  Although this may seem incongruous, the typing algorithm does not have a sufficient number of reads to unambiguously assign them to DRB3, due to the very high homology (sequence similarity) between DRB1 and DRB3.

DRB4 and DRB5 also share sequence similarity with DRB1 and may sometimes display “Insufficient data” in the typing results, although we see this result more often with DRB1 and DRB3.

To further complicate things, the DRB1 expression level is about 5X greater than DRB3/4/5 and DRB3/4/5 are not necessarily expressed in all cells.  Thus, the number of reads assigned to DRB3/4/5 is often considerably less than the number of reads assigned to DRB1, which gives the typing algorithm further difficulties in making correct, unambiguous read assignments to a specific locus.