Phylogenetic position of EBOV from the 2014 DRC outbreak

by Gytis Dudas (@evogytis)

DRC outbreak and response

In addition to the ongoing Ebola virus disease outbreak in Guinea, Sierra Leone and Liberia there is another ongoing Ebola outbreak in the Democratic Republic of Congo (DRC) which was declared on 24 September. Members of the United States Agency for International Development's (USAID) PREDICT project and DRC's Institut National de Recherche Biomédicale have uploaded 2 sequences from the DRC outbreak. These sequences are both 553 base pair fragments of the L gene, which is the polymerase. We would like to acknowledge and thank Mbala,P., Ngay,I., Makuwa,M., Wolfe,N., Schneider,B., Mulembakani,P. and Muyembe,J.-J. and Metabiota for sequencing the virus and making the sequences publicly available


There has been some confusion about the phylogenetic position of the DRC strain, despite a published report with indication of where the virus lies in the EBOV phylogeny. We decided to clarify this. We made use of the dataset we have prepared earlier for our collaboration with the Broad Institute and their partners and added a few sequences of an EBOV-like virus isolated from bats in Central Africa. We used BEAST to infer the time-scaled phylogenetic tree of EBOV sequences under previously described parameters, namely partitioning of the alignment into intergenic regions, codon positions 1+2 and 3 (giving a total of 3 partitions), with HKY+Γ(4) being used to model the substitution process in each partition under a strict molecular clock (CTMC reference prior) and skygrid as the demographic model. The MCMC chain was run for 20 million steps. We would advise caution in over-interpreting the tree that we got, the only point of this entry is to show the phylogenetic position of the virus.


This is the tree that we get:


The position of the 2014 DRC outbreak (purple) is pretty much exactly where the WHO report said it would be. In some ways this position is surprising - prior to this the EBOV tree looked a bit flu-like with what appeared to be lineage turnover within the reservoir. Now it appears that there is a lot of diversity circulating in the reservoir, which may or may not be caused by population structure of the host. The point is that we don't know - there are so few EBOV samples from bats that we have no clue about how the virus is distributed within the reservoir or how it circulates.