Phylotype consensus sequences for paper: Evidence for circulation of high-virulence HIV-1 subtype B variants in the United Kingdom

These data files contain consensus sequences of phylotypes/clusters with putative differences in transmission patterns, identified using the treestructure package on genomic data from the UK Drug Resistance Database—a central repository for routine resistance testing in the UK. These phylotypes were then used in statistical analyses of clinical markers (viral load and CD4 counts) to identify HIV-1 variants associated with increased virulence.

Files and variables

File: phylotype_consensus.fasta
Description: Unaligned FASTA file of phylotype consensus sequences for subtypes A1, B, C, and CRF02AG.

File: phylotype_consensus.csv
Variables:

• Sequence_ID: the phylotype consensus identifier.
• Collection_date: the range of dates (in years) of the individual sequences for which a phylotype was called.
• Country: the country of all the sequences within each phylotype.
• Host: the organism from which the viruses, used to generate the consensus, were isolated.
• Organism: the virus species.
• Genomic_region: the genomic region used to generate the consensus sequences.

Methods

These sequences were derived from a cluster/phylotype analysis performed on the UK Drug Resistace Database dataset for subtypes A1, B, C, and CRF02AG. Therefore, collection dates represent the range of dates of the individual sequences for which a phylotype was called. Phylotypes were identified using subtype-specific time-scaled trees estimated using treedater v0.5.3. Samples sizes of those trees were, respectively, 2714, 24100, 11331, and 2743. Phylotypes were estimated using treestructure v0.3.1 with a minimum clade size of 30 and a bootstrap support threshold of 80%. The consensus phylotype sequences were obtained using seqinr v4.2.8 with the site majority method (i.e. high-frequency character is returned as consensus) and a minimum relative frequency threshold of 80%.

Initially, some phylotypes were paraphyletic (n=30 of the submitted consensuses, listed IDs in paper Table S5; preprint available at https://doi.org/10.2139/ssrn.4929798). We resolved this by tracing the most recent common ancestor (MRCA) of all sequences in the phylotype until a monophyletic group was formed. Then called the consensus as mentioned above. There is no consensus sequence uploaded for the backbone phylotype (paraphyletic) of each subtype, which represents the large (~50% of samples) and diverse ancestral type from which all of the other phylotypes were descended.

Individual-level HIV genetic, demographic, and clinical data from the UKRDB can be accessed for collaborative projects, subject to the approval of a research proposal by the UKRDB Steering Committee.

The code used for all the analyses is freely accessible at GitHub (https://github.com/vinibfranc/HIV_UKRDB_phylotypes).