Visualising Data for Rare Disease Referrals¶
Visualising small variants¶
Visualisation of small variants can be performed using the alignment (CRAM) files. The colour and intensity of the reads, and positions within a read indicates the sequence and quality of the alignment and sequence. Reads in regions with poor mapping quality are shown in white (or lighter colours). Base quality within a read is indicated by the brightness of the colour for an individual base, with brighter colours indicating better quality.
An example of small variant visualisation is shown in Figure 8 above. In this example, reads coloured in grey indicate sequence that is concurrent with the reference sequence. Where a coloured position is shown, there is a difference from the reference sequence, with the colour denoting one of the four nucleotide bases. In the referral shown, the second patient has a single nucleotide change from the reference base (Guanine) to Cytosine. The variant position shown is present on sequencing reads in both forward and reverse directions with high base quality, which are a characteristics of high quality variants.
In assessing the quality of small variants, users should consider the quality of the alignment and the quality of the sequenced bases at the variant site. It is also helpful to assess the density of mismatched bases in surrounding regions and the alignment and base quality for these positions.
In addition to assessing variant quality, reviewing small variants is particularly useful for assessing de novo variants, to determine whether there are any reads supporting the putative de novo variant in the parental samples which may indicate potential germline mosaicism.
Visualising breakpoints for copy number and structural variants¶
Copy number variants can be visualised using both alignment (CRAM) files and coverage profiles (bigwig files). Breakpoints for copy number and structural variants can be visualised in the IGV desktop viewer or IGV.js for visual confirmation and refinement of breakpoint locations. To visualise CNVs for Rare Disease referrals, users can click on the genomic coordinates in the tabular display on the Interpretation Portal.
To review coverage for regions in and around CNVs, users can view the coverage BigWig file. This can be performed via IGV by unticking all boxes except for the BigWig file in "Available files to show" in either IGV.js or the IGV desktop viewer. Reviewing coverage profiles is particularly effective for large CNVs, for which loading the CRAM file in IGV for the entire genomic region may not be possible.
An example visualisation of coverage profiles for the region surrounding a CNV is shown in Figure 10. In this example, a loss CNV can be seen as reduction in coverage by approximately 50%, and looking at the coverage profiles for all family members can help in determining the mode of inheritance.
For additional confirmation of copy number variants and to define the breakpoints more precisely, it is beneficial to review supporting reads in the CRAM files. For large CNVs, the region is often too large to display individual reads for the whole CNV. Therefore to review breakpoints in the CRAM files, it is recommended to zoom into the each breakpoint separately.
The breakpoint regions for the example CNV in Figure 10 with visualisation of both the coverage profile and alignment are shown in Figures 11 and 12. The breakpoint can be seen as a reduction in coverage (by approximately 50%) and by the presence of "split reads" spanning the breakpoint junction. Split reads are visible as reads with a large number of mismatched bases, with all (or the majority of) reads indicating a consensus breakpoint sequence. Split reads are reads which have directly sequenced the breakpoint junction, and can be reviewed by selecting the option to show soft clipped bases. Split reads can be used for precise refinement of CNV breakpoints. To display split reads, click the cog on the right of the alignment and select "Show soft-clipped bases".
Coloured reads may also be visible at the breakpoints of CNVs. Reads are coloured according to the location and orientation of the mapped paired-reads. Deletions are typically flanked by red reads and duplications by green reads indicating a simple deletion or tandem duplication structure, respectively. Reads of other colours indicate that the CNV has a more complex structure. Note that display of red reads supporting a deletion is currently not possible in IGV.js, so using the desktop version of IGV is recommended.
Please note, that the local installation of IGV viewer has additional functionality compared with IGV.js, and allows users to review information associated with individual reads in CRAM files by hovering over the reads. However, some users have noted that visualising BigWig and BAM/CRAM files on their local copies of IGV is slow, and have found that IGV.js is sufficient for basic visualisation.
Further guidance for reviewing signatures of copy number and structural variants from CRAM files can be found https://software.broadinstitute.org/software/igv/interpreting_insert_size and https://software.broadinstitute.org/software/igv/interpreting_pair_orientations. Additional information is also described regarding review of structural variants in section 8.6.3 below.