Shifted Poisson Mixture Model (SPMM)

SPMM estimates the infection duration of a single variant or multi-variant HIV infections based on the characteristics of HIV transmission and early evolution.

SPMM takes a single fasta input file (with an extension of ".fasta") of aligned HIV envelope gene DNA sequences (either full length or a segment) obtained from one infected individual. For enhanced accuracy, users are expected to remove any recombinant sequences and APOBEC3G/F-mediated hypermutations.

SPMM's output consists of i) the pairwise Hamming distance distribution of the input HIV envelope gene sequences (grey box) along with the best fit of SPMM (red line) and ii) fitted model parameters, days since infection, the number of founder variants, Hamming distances among founder sequences, the number of sequences in each founder lineage, and the Poisson parameter (λ), along with the goodness of fit p-value, the Sum of Squared Errors (SSE) and Akaike Information Criteria (AIC).

Upload a fasta file:

Minimun Distance between Founder Variants: (1-999, default: 6)

Example: Input, Output


  1. SPMM: estimating infection duration of multi-variant HIV-1 infections, Tanzy M. T. Love, Sung Yong Park, Elena E. Giorgi, Wendy J. Mack, Alan S. Perelson and Ha Youn Lee, Bioinformatics, 32: 1308-1315 (2016). Read Here
  2. Modeling sequence evolution in acute HIV-1 infection, Lee et al., Journal of Theoretical Biology, 261:341-360 (2009). Read Here
  3. Identification and characterization of transmitted and early founder virus envelopes in primary HIV-1 infection, Keele et al., PNAS 105:7552-7557 (2008). Read Here


Supported by NIH NIAID R01-AI083115 and NIAID R01-AI095066