Trace_clip

NAME

trace_clip -- an Experiment File sequence clipper that analyses traces

SYNOPSIS

trace_clip [-w winlen_nonc] [-c cut_nonc] [-W winlen_drop] [-C cut_drop] [-s start] [-f fract_nonc] [-v] file ...

DESCRIPTION

trace_clip is used to "clip" the 3' ends of machine produced sequences. It adds a QR record to the reading's experiment file and sequence to the right of this point will be ignored for many subsequent processing steps (although note that the clipped data can be used to help find joins between contigs (see section Find Internal Joins), and to confirm single stranded regions (see section hidden-t.texi:Double stranding)). The clip position is selected by analysing the reading's traces using two simple measures. The first (nonc, or non-called over called) calculates the ratio of the area under the peak for the called base to the maximum area under each of the non-called bases at the same position. The second (drop) analyses the traces at positions where identical bases are adjacent. It calculates the ratio of the height of the lowest of the two peaks to that of the intervening trough. Both of these calculations produce values that increase 5' to 3'. The measures are averaged over windows winlen_nonc and winlen_drop that are slid rightwards from a start position start, and the windows stop when their respective cutoff values cut_nonc and cut_drop are found. The clip point is the weighted mean (using fract_nonc) of the positions at which the two windows stop.

The file arguments, of which there can be several, are processed one at a time. Each argument is assumed to be a valid Experiment File. The trace file name is read from the Experiment File; clipping is performed; and a QR identifier is appended to the Experiment File.

The default arguments are -w 51 -W 51 -c 0.3 -C 1.1 -f 0.25 -s 100.

The parameters cut_nonc, cut_drop fract_nonc can be chosen by use of scale_trace_clip See section scale_trace_clip..

OPTIONS

-w winlen_nonc

Set the length for the non-called over called window to winlen_nonc This should be an odd number.

-W winlen_drop

Set the length for the peak-to-trough window to winlen_drop This should be an odd number.

-c cut_nonc

Stop sliding the non-called over called window when the score is greater than or equal to cut_nonc.

-C cut_drop

Stop sliding the peak-to-trough window when the score is greater than or equal to cut_drop.

-f fract_nonc

Set the weight (or fraction) of the non-called over called window stop point that is to be used in the weighted mean calculation. If stop_nonc and stop_drop are the stop positions found for the two measures then the clip point is given by: clip = fract_nonc * stop_nonc + ( 1.0 - fract_nonc ) * stop_drop

-v

Enable verbose output. This outputs information on which files are currently being clipped.

-s start

Force the first window to start the calculations from position start in the sequence. This can be useful to avoid poor data at the 5' end of a sequence.

SEE ALSO

See section ExperimentFile(4).See section scale_trace_clip.