Description of the input items for the prediction server Choose the virus and type of sequence. Currently either nucleotide or protein sequence can be selected for HIV-1 or HCV. The query sequence can be pasted in this box, or Fetched from a uploaded file. Optionally, one can test the server using the example sequences. MuLDAS, by default, scan the query sequence to map the gene segments and genotyping is done gene by gene. If one of the genes is selected, mapping and genotyping are done for the selected gene only. This option tells the server how to handle the gaps in the multiple sequence alignment. "Delete all sites that include gaps" means that all the sites where gaps occur are trimmed. "Delete only gap sites of the query" means that only the sites where the query has gaps are trimmed and the gaps in the other reference sequences are treated just like an amino acid, scored as a match if two gaps occur at the same position. Multiple substitutions are corrected as implemented in the DISTMAT of EMBOSS package (http://embossgui.sourceforge.net/demo/manual/distmat.html) Multidimensional scaling can be done at the user-specified dimensionality. By default, MuLDAS operates at 10-dimensional space for both 'major' and 'nested' analyses.