Step-by-step instructions for running JOY

Before you start

JOY is an analysis and formatting program and it does NOT generate an alignment. You have to prepare your own sequence alignment. This should normally be a structure-based alignment, or at least include sequences with known 3D structure. Otherwise, you would not have been interested to use JOY.

You can use a number of programs to produce multiple sequence alignments. You can also download alignments from various databases (see HOMSTRAD). Now you have to convert your alignment to the format that JOY recognizes.

As detailed in the manual the alignment format for JOY (.ali) is similar to the NBRF/PIR format, so the easiest way to generate a .ali file is to generate a .pir file first. For example, the program CLUSTALW accepts several formats including Pearson (Fasta), GDE and GCG/MSF then allows you to save the alignment as an NBRF/PIR file. You can also use the sequence format conversion program READSEQ.

Important note

JOY uses local structural features calculated from the atomic coordinates in a PDB file. The only way to map the structural features onto the amino acid sequence is to use the sequence directly extracted from the ATOM records of the PDB file. This may not be necessarily the same as the one derived from the SEQRES records, as the model  may contain regions with no coordinates. If you are not sure, use our ATM2SEQ server to extract the amino acid sequence from your PDB file. Make sure that the sequences in your alignment are consistent with those generated by ATM2SEQ.


Suppose your .pir file looks like this. You want to include the structural information for the first protein. The structural information can be taken from the PDB entry 5p21.

  1. Decide what to call this protein. Here let's use the PDB code for the first protein, i.e, the first line of your .ali file should be '>P1;5p21'.
  2. (Skip this step if you are using the JOY server.) Get the full coordinate file for the PDB entry 5p21 and save it as '5p21.atm' in your current working directory.
  3. Edit the second line of your .ali file. If you run JOY locally, the second line can simply be 'structure'. For the JOY server, you have to give more detailed information such as this (click here for the meaning of each field):

    structureX:5p21: 1 : : 166 : :transforming protein p21/H-ras-1:Homo sapiens: 1.35: 19.6

    The easiest way to generate this line is to use the ATM2SEQ server. Simply type in the PDB code 5p21 there.

  4. The second protein in the alignment is 1hura0. You don't need the structural information for this protein. Edit the line after '>P1;1hur0' so that it becomes 'sequence'. Now your final .ali file looks like this.
  5. You can paste your alignment in the text port of the JOY server, or run JOY locally.

Created: 8 February, 1999
Last modified: 06 May 2001