journal/cge Clinical Genetics. 2018;93:936-937.siblings. He had normal growth parameters, slightly deep-set eyes, absent lingual frenula and 9/9 Beighton joint hypermobility. Skin was hyperextensible in the neck, eyelids, distal forearm, dorsum the hands and abdomen. Large papyraceous hemosiderotic scars were noted on the knees and elbows, and to a lesser extent on the dorsum of feet (Figure 1). Depressed and linear scars were also noted on the forehead and supraorbital region. He had normal echocardiogram, prolonged Partial thromboplastin time (PTT) (51.3; normal <26) and reduced factor XI level (46; normal >75). A clinical diagnosis of cEDS was made and the patient and his parents were recruited with informed consent and the study was approved by KFSHRC (IRB, RAC#2121053). Whole exome sequencing (WES) in the index revealed a homozygous indel in COL5A1 (NM_001278074.1: c.5072_5077delinsT;NP_001265003.1:p.Lys1691Ilefs*13). This was confirmed by Sanger sequencing; the parents and one sibling were heterozygous carriers and the other sibling was non-carrier (Figure 1). Carriers lacked suggestive manifestations of cEDS, including hypermobility, consistent with the recessive nature of this exon 64b variant, which was absent in gnomAD as well as 950 ethnically matched exomes. No candidate variants were identified in other connective tissue genes and this was the only novel homozygous coding or splicing variant identified by WES in the index.While the occurrence of recessive variants in genes known to cause human diseases strictly through dominant variants is well known, this is typically due to the difference between loss of function mechanism in recessive variants vs dominant negative or novel gain of function mechanisms in dominant variants. 4 The mechanism behind the recessive nature of the variant we describe in COL5A1 appears unique in that while similarly loss of function like all reported dominant COL5A1 variants, this is the only variant to date that affects only 1 of the 2 COL5A1 transcripts. The reported dynamic difference in abundance between these 2 transcripts suggests that the apparent sparing of isoform a allowed the carrier parents to compensate, especially since isoform b only accounts for <30% of COL5A1 in skin and knees. 3 It will be of interest to observe the clinical consequence of pathogenic variants in exon 64a in humans and their mode of inheritance.