The COLLECT clause specifies what the data the statistics are collected over and, indirectly, how the statistics are tagged. There are three options permitted.
The ALLTAG option mean that statistics are to be collected for every data point irrespective of tag. The tag assigned to a particular statistic is the same as the `stat-option' (except of course for ALL).
The EACHTAG option specifies that statistics are to be collected individually for each tag that appears. The tag for these statistics is the tag of the data point followed by a colon followed by the `stat-option'.
The final option allows one to select a set of tags for which data is collected and do statistics on that set. The tags are specified using tag patterns which allow the same wild cards as atom or cell selections, see Atom Selection. The tag used for this option is the same as the name specified in the command, and therefore, specifying more than one `stat-option' will result in an ambiguous table. The use of EXCEPT with this option allows one to collect data over all tags except those specified. This option may be used for collecting statistics over the backbone or sidechain of a protein.
The range of data over which the statistics are collected is determined by the PLACE option described in the next node.