Define Regions

Reference sequences can be divided into (possibly overlapping) regions, each of which is a contiguous range of positions in one sequence.

How to define regions using coordinates or primers

How to define regions using coordinates

A region can be defined by specifying its first and last coordinates, with respect to the numbering of positions in the reference sequence. SEISMIC-RNA uses the following conventions:

  • The first (i.e. 5’) position in the reference sequence is numbered 1, as is conventional for biological sequences. Consequently, the last (i.e. 3’) position in the reference sequence is assigned the number equal to the length of the reference sequence.

  • A region includes both the first and last coordinates. Note that the length of the region therefore equals the last minus the first coordinate plus one. For example, if the reference sequence is 10 nt long, then the region whose first/last coordinates are 4 and 8, respectively, will include positions 4, 5, 6, 7, and 8; but not positions 1, 2, 3, 9, or 10.

  • The first coordinate must be a positive integer, and the last must be an integer greater than or equal to the first coordinate minus one and less than or equal to the length of the reference sequence.

  • Setting the last coordinate to the first coordinate minus one creates a zero-length region; this behavior is permitted in order to handle this edge case smoothly, but rarely if ever has practical use.

On the command line, define a region using coordinates via the option --coords (-c) followed by the name of the reference and the two coordinates. For example, -c refA 34 71 would define a region spanning positions 34 to 71 of reference “refA”.

How to define regions using primer sequences

For samples that were prepared as amplicons (using RT-PCR with a pair of primers flanking a specific sequence), it is usually easier to define regions using the sequences of the forward and reverse primers compared to using coordinates. SEISMIC-RNA will compute the coordinates from the primer sequences using the following conventions:

  • The entire sequence of the forward primer and the reverse complement of the reverse primer must both occur exactly once in the reference sequence (no mismatches or gaps are permitted).

  • Primers will cover up any mutations during RT-PCR, so the sequences they bind provide no information for mutational profiling. Thus, SEISMIC-RNA makes the region start one position downstream of the 3’ end of the forward primer, and end one position upstream of the 5’ end of the reverse complement of the reverse primer.

  • Artifacts may occur near the ends of the primers. If so, then a number of positions at the both ends of the region can be ignored using the option --primer-gap {n}, where {n} is the number of positions to ignore.

On the command line, define a region using primers via the option --primers (-p) followed by the name of the reference and the two primer sequences; the reverse primer must be written as the sequence of the oligonucleotide itself, not its reverse complement. For example, if the sequence of “refA” is TTTCGCTATGTGTTAC, then -p refA TCG AAC would define a region, depending on the primer gap, spanning these positions:

  • 6-11 (CTATGT) with --primer-gap 0 (the default)

  • 7-10 (TATG) with --primer-gap 1

  • 8-9 (AT) with --primer-gap 2

How to define regions using a file of coordinates and/or primers

To create many regions at once, and to do so reproducibly, it is more convenient to define them in a file than on the command line. For information on this file, see Metadata for Regions. Provide a file defining regions using the option --regions-file (-s), for example -s regions.csv.

How to define multiple regions simultaneously

How to define multiple regions for one reference

On the command line, the -c and -p options can be given multiple times to define more than one region for a reference. In a regions file (-s), each reference can be given multiple times, as long as the name of each region of that reference is unique.

For example, typing -c refA 34 71 -c refA 56 89 -c refA 103 148 on the command would make three regions for reference “refA”:

  • positions 34 to 71

  • positions 56 to 89

  • positions 103 to 148

Note that regions are allowed to overlap or contain each other, as with 34-71 and 56-89.

How to define regions for multiple references

Regions are defined for each reference sequence separately and have no effect on regions of any other reference sequence. For example, suppose that for three references, “refA”, “refB”, and “refC”, one region is defined for “refA” and one for “refB” using the option -c refA 34 71 -c refB 86 130.

  • “refA”: positions 34 to 71

  • “refB”: positions 86 to 130

References “refB” and “refC” do not get region 34-71, nor do “refA” and “refC” get region 86-130. If no regions are defined for a reference sequence (here, “refC”), then one region, spanning the full reference sequence and named “full”, is created automatically. This applies to regions given as coordinates or primers on the command line (with -c and -p) as well as in a file (with -s).

How to use the full length of a reference sequence automatically

To use the full length of a reference sequence, simply avoid specifying any coordinates or primers for that reference on the command line or in a regions file. Keep in mind that you may specify coordinates or primers for any other references, since all references are split into regions separately.

How to name regions

Regions are named as follows:

  • If the region is defined in a file, then its name is taken from the “Region” column of the file (see Metadata for Regions).

  • If the region is defined on the command line by its coordinates or primers, then its name is the first and last coordinates, hyphenated (e.g. -c refA 34 71 would create a region named “34-71”).

  • If the region is created automatically because no other regions were defined for its reference sequence, then its name is “full”.