By default, all reads in the inputs are written to the output SFF file. If the “-i” (Include only these reads in output) or “-e” (Exclude these reads from output) options are used, the accession numbers given in the files specified will change the reads that are output. (If both -i and -e are used, the output will include the reads that are in the -i file and not in the -e file.) Those files should list the accessions one per line, where the accession should be the first word on the line after an optional ‘>’ (i.e., if the line begins with ‘>’, that character is skipped, and the following characters up to the first whitespace character is read as the accession). These options are cumulative, i.e. if multiple -i options are given, the reads included will be the union of the -i accession lists.

These two options adjust the trim points for some or all reads in the output SFF file. The specified “trimfile” should contain one or more lines consisting of (1) a read accession number, (2) a starting trimpoint, and (3) an ending trimpoint, separated by whitespace characters or where the trimpoints are separated by a dash (e.g., “accno 12 543” or “accno 12-543”). The trimpoint values are 1-based positions that denote the first and last base of the trimmed region (e.g. for a read 800 bases in length, the lines above specify that bases 1-11 and 544-800 should be ignored, and bases 12-543 form the trimmed region of the read). A value of 0 specifies that the beginning or end of the read should be used (e.g. for a read 800 bases in length, the line “accno 12 0” sets the trimmed region to 12-800).

This option “splits” the input reads into separate output files, based on the multiplexing sequences occurring at the beginning of the reads. The argument to the option is an MID set name (occurring in the MID configuration file). Each input read is matched against each MID in the set, and output to a separate file. The output files will be named by adding the MID name as a suffix to the output file (but before the “.sff” suffix). So, if the default GS Multiplex Identifiers (MID) Kit set is given as the set name and the output filename is “454Reads.sff”, the output files will be “454Reads.MID1.sff”, “454Reads.MID2.sff”, and so on. Note: If no reads match a given MID sequence, the corresponding output file is not written (i.e., only files with one or more reads will be output).

[-mcf filename]

This option tells sfffile to generate an output file containing a certain number of reads, by randomly “picking” reads from the input. The argument number can be followed by a ‘k’ or ‘m’ to specify thousands or millions of reads, respectively. Reads in any input SFF files are merged directly into the single output SFF file generated by this command. If SFF files are not present in a Data Processing folder (e.g. for a Run whose data has been processed with a version of the 454 Sequencing System software anterior to 1.0.52), the signal processing step of the GS Run Processor application must be rerun on the Data Processing folder, and can then merged into the output SFF file generated by the sfffile command. Input “D_...” directories (with data generated on a GS FLX+ Instrument) may be prepended with a list of regions separated by a colon. For example, “1,3-5,7:R_dir/D_dir” tells sfffile to use regions 1, 3, 4, 5 and 7 of R_dir/D_dir. An optional multiplexing information string can be prepended to each file/data-directory argument.

Path to one or more SFF files and/or Data Processing directories (“D_...”). Reads in any input SFF files or the SFF files in the Data Processing directories are merged directly into the single output SFF file generated by this command. Input “D_...” directories (with data generated on a GS FLX+ Instrument) may be prepended with a list of regions separated by a colon. For example, “1,3-5,7:R_dir/D_dir” tells sfffile to use regions 1, 3, 4, 5 and 7 of R_dir/D_dir. An optional multiplexing information string can be prepended to each file/data-directory argument (see Appendix 4.7).