NOTE: ADRA scores the types of changes seen in HIV in response to anti-retroviral therapy; it is intended for research use only.
The HIV Resistance Database consists of a compilation of mutations in HIV genes that confer resistance to anti-HIV drugs. The database is based on the review article, Mutations in Retroviral Genes Associated with Drug Resistance. An annually updated publication appears in the HIV Database Sequence Compendium. It allows for user specified searches on information pertaining to drug-resistance conferring mutations. From the database, links lead to Medline abstracts of publications from which the data was obtained. ADRA uses this database as a library for analysis. By aligning a query sequence to any provided or self-supplied referent sequence, differences between the query and referent sequences are highlighted and searched in the HIV Resistance Database, if referenced as being drug-resistance conferring.
ADRA is designed to identify possible anti-viral drug resistance-associated changes in protease, reverse transcriptase, integrase and envelope sequences. The program accepts FASTA formatted sequences commonly generated by sequencing (for example by the Perkin-Elmer and Visible Genetics HIV resistance typing kits). The program will also accept amino acid (aa) sequences. In the case of nucleic acid sequences, ADRA translates all three sense frames to aa. These aa sequences are then matched to a library of sequences using BLAST to identify the gene region(s).
The use of a library of sequences allows a wider range of comparisons; as any one sequence or consensus may not necessarily provide a good overview of what may be considered a novel change in the query sequence. Hence using a large library of different sequences will be more beneficial for the analysis. It is important to note that the HIV Resistance Database does not necessarily contain only resistance mutations relative to HXB2r. Results may vary depending on the referent sequence. In addition users can supply their own referent sequence(s), which may be the optimal reference sequence to determine drug sensitivity or resistance; an example of when this might be useful is during pre- or post therapy.
The product of the BLAST search is an alignment which delineates the gene regions, i.e. protease, RT or integrase, with markers. Amino acids which mismatch the selected referent sequence are highlighted and treated as mutations, ambiguous characters are written in red. These highlighted changes are then submitted to a library which contains each resistance-conferring mutation recorded in the HIV Resistance Database at Los Alamos. If matches are found, these mutations are listed in a table, with a summary of their database record and a hyperlink to the more detailed record in the HIV Resistance Database.
The program also detects stop codons, the position (co- ordinates) of the sequence relative to the referent and the correct reading frame. These results are also presented in a table. The frequency and position of ambiguous characters may provide important information about the evolving viral diversity within the host, giving insight into the possibility that the patient may be in the process of becoming sensitive or resistant to one or more of the drugs he or she is being treated with. The presence of stop codons quickly tells the user that the sequence is a pseudogene or includes a frameshift.
In addition to analyzing one sequence, an alignment may be provided containing several samples, possibly from the same patient. More conclusive results about evolving resistance to current anti-viral drug therapies could be found through this option, since a wider sampling of the patient's viral population can be made in this way.
The program can only analyse sense sequences and will not correct sequences with frameshifts. The program makes no attempt to integrate multiple mutations for resistance to generate an overall score for resistance (for example, the RT positions 41, 67, 70, 210 and 215 associated with Zidovudine/AZT resistance, are scored independently). Although the program links into the HIV resistance database, no editorial interpretation of the differing types of resistance tests in common use in the literature is made. Hence a resistance entry cited as 20 fold for one drug, may differ from an entry of 20 fold for another. The user should investigate the primary citation referred to by the hyperlink for details. Cross resistance data is also presented where available, once again, the precise nature of this data should be investigated in the primary citations.
Submitted sequences and results are held on the Los Alamos server for no more than 24 hours after submission and are then automatically deleted. However the web connection is NOT secure and it is the users responsibility to ensure that the submitted sequence does not contain any case numbers, patient details or other identifiers which could breach confidentiality!
This software is made available on a best effort basis and although we have made every effort to ensure that the hyperlinks found for any sequence are accurate, their completeness and accuracy will ultimately be defined by the quality of the submitted sequence. We therefore can accept no responsibility for any omissions or errors made by the program on real world data.
You are advised to use this program as companion to other methods, not a replacement, and to carefully inspect the citations found by the hyperlinks to check that the very simple interpretation shown in the summary table is adequate.