For an assembly containing n reads, the file reads.ids will contain n+1 lines. The first line contains the number n (useful information for memory-allocation routines such as READ in System). The following n lines contain the read names. Note that the ordering of these names is extremely important because it implies the read ids: that is, the first name (on line 2) is for read 0; the second name is for read 1; and so forth. A sample file, with 5 reads, might look like this:
5 read_id_0 read_id_1 read_id_2 read_id_3 read_id_4
If necessary, it is easy to create reads.ids from reads.fasta using UNIX commands. First, run the following command:
grep ">" reads.fasta | cut -c2- > reads.ids
This gets all the read names in the correct order. Next, you must add the first line to reads.ids, which is easily done manually. Run wc -l reads.ids to find out the number of reads (i.e., lines in the file) and prepend it to reads.ids.