Gene Codes Sequencher 4.8 | Quality Scores Tutorial

Quality Scores

The process of generating DNA sequencing data involves many steps. Part of the process involves recording the intensity of fluorescence as a nucleotide-associated dye passes a laser. A subsequent step requires analysis software to convert these recordings into base calls. The ability of base calling software to accurately interpret raw traces varies with the quality of the data. The most common early base callers called an ambiguous base, or an “N,” whenever the quality of the raw data failed to yield a clear signal for A, C, G, or T. Advances in base callers led to the evolution of the quality or confidence score.

Quality scores are numeric values associated with each base call that define how likely it is that a base call is incorrect. The most common scale is from 1-60, where “60” represents a 1/10⁶ chance of a wrong call,“50” a 1/10⁵ chance, 40 a 1/10⁴ chance, etc. Depending on the program used to generate the confidence value, the quality score may be based on peak height, the presence of more than one peak, and/or the spacing between the peaks.

Many instruments sold today come with base callers that generate confidence scores as an integral part of the chromatogram file. These include Applied Biosystems’s KB base caller, MJ Research’s SCF files, and the ESD files generated by the Beckman CEQ and the Amersham ⁄ GE MegaBace. Some labs use Phred, TraceTuner, or other third-party programs to generate confidence values in separate documents that are associated with the chromatogram files. Various Base Calling software programs generate a variety of different file formats. To import sequences into Sequencher with all of the associated data as one file, you may import a PHD file or you can import the individual sequence ⁄ qual ⁄ trace files. The key to associating the data within Sequencher is the naming of the file. If you have any specific difficulties with your file format, please feel free to ask Gene Codes Technical Support for assistance.

Sequencher is capable of working with confidence values from all of the above files types. Sequencher imports and displays confidence ranges in three user-defined ranges: low, medium, and high. Sequencher trims data based on confidence scores, and automatically navigates to regions of low confidence score. You can view summary confidence information about your sequences in the Project window, the Sequence Editor, and in the Get Info box.

Sequencher displays the confidence scores in three user-defined colors so that display of confidence will not interfere with your display of base calls.

Quality Scores

Launch Sequencher.
From the File menu, choose Import > Sequencher Project… and load QualScore Project.SPF.

Sequencher will load nine auto-seq fragments into your Project window. The raw data was base called with Applied Biosystems’ analysis version KB1.1, which generates confidence scores for each base call. Next to the name of each of the ABI sequences loaded into your Project window is a value for % Quality. This number defines the percent of quality bases for each of the sequences.

Under the Window menu, select User Preferences….
One of the options in the General category is called Confidence. Select that topic.
Set the confidence ranges as shown in this figure:

With these settings, base calls with scores between 0 and 20 will be displayed with a cobalt blue background, those with scores between 21 and 39 will be displayed in cadet blue, and those 40 and above will be displayed in aqua. Bases that you edit by hand (and are therefore sure of) will be displayed on a white background.

Exit User Preferences by closing the window.
From the bottom of the View menu, make sure that Display Base Confidences is checked.
From the Select menu choose Select None.
Double-click on any sequence to open its Sequence Editor window.
Select a range of bases that includes some low quality bases.

You can see the lower quality bases in darker colors. Note the number of low quality bases in your selection appears above the bases in the sequence editor.

The quality scores can be used to trim the data before assembly.

Close the Sequence Editor window.
Select all nine sequences by choosing Select All from the Select menu.
From the Sequence menu, select Trim Ends…
Click on the Change Trim Criteria button at the top of the window.
Deselect all options with the exception of the following:

For trimming at the 5’ end, check the box and insert settings as shown, so that less than three bases will be allowed in a 25 base window with quality score below 25.

Do the same Trimming for the 3’ end.

Select the option in the Post Fix portion of the window to remove leading and trailing ambiguous bases.

Click OK in the Ends Trimming Criteria window.
Click on the Trim Checked Items button.
When you are asked to confirm your decision, click Trim.
Close the Ends Trimming window to return to the Project window.

You are now ready to view the data in an assembly.

Select all 9 items in the Project window. (They should be selected already.)
Click on the Assemble Automatically button at the top of the Project window. When the assembly completes, click Close.
Double-click on the Contig icon to see the Overview of the sequence assembly.
Click on the Bases button to see the aligned sequences.
From the Select menu, choose Next Contig Disagree to find the first discrepancy between the overlapping fragments.

You are probably looking at base 126 where one, low-quality base call is a ‘T’ but the overlapping fragments all have a medium quality ‘C’ at that position.

To review the trace data, click the Show Chromatograms button at the top of the window.

All of the trace data have evidence of both the ‘C’ and ‘T’ base. This is a location where there is a heterozygous base, both a ‘C’ and a ’T’, or a ‘Y’. In general, most base callers assign low quality scores to heterozygotes.

With the consensus base still selected, type a ‘Y’.

Not all bases that require editing will be flagged with a disagreement. For example, mixed bases that may have been called an “N” in a previous base caller will now be called a low quality A, C, G, or T. Also, consensus sequence based on single coverage will never create a disagreement unless you have a Reference Sequence. In order to examine the contig with greater scrutiny, you may want to look at all of the low confidence bases using the Select > Next Low Confidence Base menu command.

The Select > Next Ambiguous Base menu item is also sensitive to the presence of confidence scores. This feature enables you to navigate from one position in the consensus where the base call is ambiguous to the next. Sequencher defines the following as ambiguous consensus bases:

Where there is a disagreement between contributing bases
Where one of the contributing bases is ambiguous
Where there is a gap in the contributing sequence
Where the quality of the only sequence(s) contributing to the consensus falls in the low range as defined in User Preferences

Close the project without saving and exit Sequencher if desired.