Known Issues¶
Current Issues¶
These issues are those that currently affect the cDSS
| Issue ID | Description | Impact | Workaround |
|---|---|---|---|
| IPDSS-11830 | When viewing the interactive plots for cases with very high ploidy, the values on the tumour coverage plot copy number axis label may become squished together and difficult to make out. | We believe the impact will be minimal to users due to the rarity of these cases (at the time of this analysis only ~5% of production cases have a ploidy of >4 and therefore may be affected by this issue). | Manually zooming in using the browser zoom can make the numbers visible. No other workarounds are available at present, however, we are investigating a fix. |
| IPDSS-11891 | The variant details page displays the incorrect tooltip explainer for paired reads and supporting reads per billion. The tooltip displayed for both values is "Split reads support for a structural variant called" which applies to split reads only. | This could cause user misinterpretation if the Clinical Scientist was not familiar with these values. | Explainations for these values are available in the cDSS user guide. |
| IPDSS-12218 | In the GTAB report page, "Diagnosis date" is shown under the "Tumour and sample details" section, however when the PDF report is created this data point appears under the "Patient information" section. | We believe the impact is minimal to users as the Diagnosis date is available on both the GTAB report page and PDF report. | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| IPDSS-12620 | It is not possible to search the variant grid using Ensembl identifiers, despite being listed as a valid input in the search bar. If users wish to search through un-domained/un-tiered variants that do not have a gene name, for example non-coding genes, no results are returned even if a variant is found in that region. | The impact is believed to be minimal as the use case for searching through un-domained/un-tiered variants that do not have a gene name is likely to be small and these genes are unlikely to have a significant clinical impact. | Use the gene name instead of the Ensembl identifier where possible, if the gene does not have a gene name, search by the variant coordinates for your gene instead. |
| IPDSS-12705 | In the case summary, the sample and sequencing quality information for germline samples does not contain headers that explain the source of the data (referral vs WGS). Germline sample quality information is derived from the referral, while germline sequencing quality information is derived from the WGS data. | Ths could lead to user misinterpretation regarding the source of the data displayed (WGS vs the referral). | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| IPDSS-12730 | Once a case is closed in the cDSS, users are blocked from adding further information (classifications, comments etc) and any input boxes are greyed out. For the GTAB page, the notes, outcome, recommendation and attendees fields are still available for input, however if the user adds text here it cannot be saved. | This prevents the GTAB page being populated once a case has been closed. The current design could cause potential data loss if the user was unaware their inputs did not save. | For all cases with 'Closed' status, a 'Case closed - actions disabled' alert is visible at the top of all pages including GTAB. The case status can be updated if users wish to add information to the GTAB report. |
| IPDSS-13037 | When the CNV plots are zoomed in to a single basepair resolution, the total allele count and minor allele count lines extend 1bp beyond the copy number data shown in the tumour and germline absolute allele count plots. | If the user was to zoom in at such high resolution on the plots, this may cause user misinterpretation as to why the total allele count and minor allele count lines continue for 1bp more than the observed copy number data. The observed copy number data covers the region specified in the variant coordinates. | These regions can be reviewed in IGV. |
| IPDSS-13047 | Where the cancer pipeline uses multiple variant callers for detection of particular variants with increased sensitivity (e.g., ITD variants in FLT3), the cDSS will display variant calls from all variant callers (Manta and PINDEL), whereas the HTML report will only report on the most accurate variant call (PINDEL). | The cDSS may display all additional calls that are absent from the HTML report which may cause user interpretation. The impact should be minor in terms of affected cases. | The HTML report can be referred to in these instances as redundant calls are filtered out. Additionally the region can be reviewed in IGV. |
| IPDSS-13068 | In the interpretation drawer, we show the size of the variant in bp but incorrectly display kb as the units. | This could be confusing if the user was to view the size in the interpretation drawer and variant grid, which shows the correct value in kb. | Refer to the variant grid for the correct variant size. |
| IPDSS-13091 | The cDSS should not display calculated tumour content when the tumour sample has a high percentage of somatic variants (>40%) with a low VAF (<6%) as tumour content cannot be reliably estimated for these cases, however the value is displayed. | There is a risk of user misinterpretation due to us displaying a potentially misleading calculated tumour content for cases with low VAF. | We display a case disclaimer to warn users that a case has low VAF and to interpret results with additional caution. This warning is absent from the GTAB summary page (and the corresponding PDF version) therefore we recommend this information is manually added as a comment to the GTAB page. |
| IPDSS-13283 | In the germline CNV grid, the tooltip displayed for the confidence value assigned to CNVs is incorrect and should say say "Copy number variant with FILTER=PASS as determined by Dragen's CNV calling software" instead of "Copy number variant with FILTER=PASS as determined by Dragen's somatic CNV callign software". | The use of the term "somatic" is inappropriate for the germline CNVs, as Dragen performs somatic and germline CNV calling and could lead to user misinterpretation. | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| IPDSS-13368 | The coverage plots may show CNV calls that appear/disappear depending on the zoom level. | This could cause user misinterpretation and delay to care as more time is needed to confirm whether these calls are real or artefacts. | Clinical scientists should cross reference these calls with the information shown for CNV calls in the somatic or germline CNV variant grid and IGV. |
| IPDSS-13432 | If a germline small variant has more than one ClinVar record, only one is displayed on the germline small variant grid. There is no logic as to which entry is displayed and hidden from the grid. | This could lead to user misinterpretation and even false negatives or positives if the Clinical Scientist does not see the additional entries on the variant details page or in ClinVar. | All entries are visible from the variant details page and externally on the ClinVar database. |
| IPDSS-13504 | The tumour BAF plot legend is outdated for cases ingested following plot updates. From release 6.11.0 onwards, the tumour BAF plot legend should no longer include "Observed data is plotted as box showing median (dark line) and interquartile range (lighter box) for the major and minor allele frequency". | This could lead to user misinterpretation regarding how the plots are populated with the raw data. | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| IPDSS-13609 | The germline BAF plot legend is outdated for cases ingested following plot updates. From release 6.11.0 onwards, the germline BAF plot legend should no longer include "Observed data is plotted as box showing median (dark line) and interquartile range (lighter box) for the major and minor allele frequency". | This could lead to user misinterpretation regarding how the plots are populated with the raw data. | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| IPDSS-13610 | The germline BAF plot legend should not reference "Expected data is shown as a black line" as this legend is not applicable to the plot. | This could lead to user misinterpretation regarding how the plots are populated with the raw data. | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| IPDSS-13708 | Tumour first cases ingested post Dragen v4 release still display an alert regarding the BAF plot improvements pending. This alert should not be present as the improvements were included as part of the Dragen v4 release. | This alert being present could reduce user confidence in the BAF plot where there is no need to be. | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| IPDSS-13753 | When adding a variant to the GTAB form from the variant grid, if the GTAB form is opened in another tab and the variant is removed from the GTAB form, the first tab will not be updated unless the page is refreshed. If the user clicks on the 'Remove from GTAB' button, the error message 'You have reached the maximum of 50 variants that can be exported. You can add more variants by removing existing ones.' is shown which does not reflect the actual error. | We estimate the probability and impact of this issue to be low as it requires two tabs to be open and the user not refreshing the variant grid tab once changes have been made to the GTAB form tab. | We recommend working from a single tab when adding and removing variants from the GTAB form or refresh the variant grid page if changes are made to which variants have been added to/removed from the GTAB form. |
| IPDSS-13772 | For cases where the global analysis plots are unavailable, the Case Summary page still displays a link to 'View all global analysis plots'. | This may be confusing to the user where there are no global analysis plots available. This may cause user misinterpretation and delay to care due to time lost trying to locate the global analysis plots elsewhere (such as the HTML, where they will also be absent). | On the app sidebar, the global analysis plot icon is greyed out, and in the Case Summary page, where the Circos plot is usually displayed, an error message reads 'Global analysis plots unavailable'. |
| IPDSS-13817 | If a variant has more than one gene annotation, and the display gene (belonging to the highest domain/tier) has only one transcript, even though the additional gene annotations have more than one transcript, the user is not able to view these additional transcripts for the additional genes from the variant grid overlay. | The impact is expected to be minimal due to the display gene being the most clinically relevant in the majority of scenarios. | Transcript details for all gene annotations are available on the variant details page. |
| IPDSS-8857 | When a SV does not have any associated paired or split read value, due to the value not being provided by the variant caller, the cDSS displays the value as 'Not applicable'. | This could be difficult to interpret by the user leading to user misinterpretation. | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| IPDSS-9981 | Users are unable to see the 'IGK' gene on the genetrack for SVs and CNVs involving this gene. Normally when a SV or CNV is called, the user can search for the gene or click on the variant coordinates and the plots will zoom into the region and show the overlapping genes on the gene track. For variants in IGK, the gene is absent from the gene track so the user cannot see what portion of the call the gene overlaps with. | This could cause user misinterpretation and delay to care due to needing to validate the plots in other systems. | The variant can be viewed in IGV where a gene track is available. |
| PDSS-1268 | Cancer Hotspot annotations are not yet available for splice variants or Indels, and we are not utilising the latest version of the cancer hotspots data (updated February 2026). | User misinterpretation may occur if users are unaware that indels and splice variants will not receive cancer hotspots annotations from the cDSS even if they are present in the cancer hotspots database. This could cause false negative reporting if the user did not verify the data externally. | Until we have updated our cancer hotspot annotations to include indels, splice variants and the latest hotspots, please nagivate to cancerhotspots.org to check for the presence of your variant in the database. |
| PDSS-2335 | The total variant counts are discrepant in cDSS and HTML. The HTML and cDSS provide summary variant counts for SNVs, SVs and CNVs, however these values are discrepant between the HTML and cDSS due to the HTML report performing filtering on poor quality variants, while the cDSS utilises a raw variant count. For SVs, double counting is present in the HTML which does not occur in cDSS. | If users were to compare these numbers from the HTML and cDSS, they may be initially confused by the difference, be unsure which is correct (and which to base their interpretation on) and have reduced confidence in the validity of the data presented in either format. | At present there is no workaround, other than documenting this in the user guide for user awareness until a fix is released. |
| PCA-6479 | Variant filtering for reported variants in tumour first cases in the cDSS may differ from that in HTMLs: in these cases, we apply both somatic and germline prioritisation algorithms to all variants with GEL internal population germline allele frequency under 2%, so all such variants will appear on the cDSS. In the HTMLs and CSV tables, all variants with either GEL internal or gnomAD population germline allele frequency over 1% and no potential effect on cancer predisposition genes (nothing in the "Interpretation as germline" field) are filtered out. | This bug may cause user misinterpretation due to the discrepancies between the HTML and cDSS, leading to delay to care if users spend additional time investigating these discrepancies. | Users can refer to the HTML for analyis of tumour first cases until a fix is released. |
Please report any bugs or issues with the DSS via the Genomics England Service Desk.
Resolved Issues¶
Resolved Issues
These issues are those that existed previously but have been resolved in this version of the product. Where an issue affects cases ingested during a certain time period, the issue will remain on this list for user awareness.
| Issue ID | Fix Version | Description | Impact | Workaround |
|---|---|---|---|---|
| IPDSS-13962 | 6.19.0 | When filtering the SV and CNV grid by supporting reads (split and/or paired reads) or copy number, the cDSS does not handle float values correctly. The user should only be able to accept integers, however if a float is input, the cDSS will not produce an error. These data points cannot be filtered on non-integer values therefore the filtering will not work as expected. | The impact is expected to be minimal as it's unlikely users will filter these data points using float values, as it does not make sense. | A fix has been released for this behaviour therefore this bug is no longer present in the cDSS. |
| IPDSS-13963 | 6.19.0 | Search bar queries do not zoom into the region of interest in plots, they remain at the whole genome view. In previous versions of the cDSS, the plots zoomed into the region of interest submitted by the user via the searchbar. | This change in behaviour when using the search functionality and the plots can create user misinterpretation and lead to delay to care due to additional time spent needing to manually zoom into the region of interest when using the plots and verifying the correct region is visualised. | A fix has been released for this behaviour therefore this bug is no longer present in the cDSS. |
| IPDSS-13964 | 6.19.0 | When the size filter is applied to the SVs and/or CNVs, it is not possible to remove input from the minimum value. | This behaviour could cause user confusion when applying the size filter to SVs and/or CNVs. | A fix has been released for this behaviour therefore this bug is no longer present in the cDSS. |
| IPDSS-13969 | 6.19.0 | When clicking on the 1-2kb region shown in the variant grid for DUX4 rearrangements, the tumour SV plot zooms into the start of the region instead of showing the entire region. DUX4 rearrangements are an exception to how other SV breakpoints are displayed in the cDSS. Usually we show the precise breakpoint coordinates and zoom into that specific location with certain padding when looking at the plots. Whereas for DUX4 rearrangements we show a region instead of precise breakpoint coordinates. | Treating the visualisation of DUX4 rearrangements as other rearrangements can cause user misinterpretation, a delay to care and in the worst case scenario if the user was not aware of this behaviour with the plots, a false negative. | The region of the partner break-end location shown in the variant grid and it is possible to adjust the view manually in the cDSS. Alternatively the HTML and CRAM files for view in IGV is available as alternative viewing options. |
| IPDSS-13961 | 6.19.0 | When modifying the input of filters that take numerical values, such as VAF, if the user removes numbers after a decimal place the cDSS will automatically reset the value to 0. | This bug causes no risk to the user and is a pain point when modifying numerical values for filters, requiring the user to input their value again. | A fix has been released for this behaviour therefore this bug is no longer present in the cDSS. |
| IPDSS-13049 | 6.13.0 | The Gene mode of action annotation following NGIS Petra release and NGIS Quasar release, H3C3, RXRA, SOS1, YAP1, RASA1, RICTOR, RAD51, and ROCK1 genes were not be labelled in the DSS with a mode of action or flagged as being present in CGC when compared with the HTML report due to a version difference. Additionally, following NGIS Quasar release, SET remained labelled as an oncogene in the DSS. Whereas, SET mode of action is N/A in the HTML report. | This could cause user misinterpretation due to the discrepancies in the cDSS and HTML, delay to care and in the worst case scenario potential false positive and false negative results if cDSS data was used. | These gene mode of action discrepancies between the DSS and the HTML report have been fixed as of the 6.13.0 release (22/10/2025) and any cases arriving to the DSS from this point onwards will be unaffected. Please check the mode of action of these genes in the HTML report if the case arrived to the DSS between NGIS Petra release (11/06/2025) or NGIS Quasar release (27/08/2025) and DSS 6.13.0 release (22/10/2025). |
| PDSS-1084 | 6.2.0 | We have identified a bug in our plot generation, where reported sex is being used rather than genomic sex. If a cases genomic sex differs from the reported sex, the "expected coverage" line across chrX would display incorrectly. For further information and an example, see the "Tumour Copy Number Variant Plots" section in Visualisations. | This could cause user misinterpretation when visualising the plots and cause delay to care due to time spent investigating the issue. | A fix has been released, however for cases ingested since cDSS go live and the fix (6.2.0) user awareness of the issue when visualising the plots should reduce the likelihood of user misinterpretation and delay to care. |
| PDSS-2213 | 6.7.0 | Cases processed by Dragen v4 (NGIS release Orion) are missing the following case disclaimer in the cDSS. "The overall ploidy for this sample is estimated to be |
The absence of this warning for cases ingested while the bug was active may lead to user misinterpretation regarding the limitations of the Dragen v4 pipeline. | For cases ingested prior to the release of the fix, there is no workaround other than documenting the absence of this warning in the user guide for user awareness. |
| PDSS-1937 | 6.4.1 | At the time of the ‘Nembus’ release on the 9th October, 6 new loci were added used to annotate additional upstream/downstream variant data for cancer cases. These new loci were not included in cancer DSS. These variants are still visible and reported in the correct domain in cDSS, but they are not specifically annotated as being associated with the 6 new loci, which can still be visualised in IGV or the HTML report. The bug has been fixed as of the 6.4.1 release (13/11/2024) and any cases arriving to the cDSS from this point onwards will be unaffected. | This bug could lead to user misinterpretation, delay to care and false negative results due to the loci not being displayed correctly in cDSS. | If you are interested in these loci and the case arrived to the cDSS between 9/10/2024 and 13/11/2024, please check the HTML report and IGV. |
Cancer loci for PDSS-1937
| Loci | chr | start | end |
|---|---|---|---|
| TAL1_upstream | 1 | 47232226 | 47242255 |
| TLX3_upstream | 5 | 171205524 | 171265681 |
| TLX3_downstream | 5 | 171312140 | 171349100 |
| ABL1_upstream | 9 | 130708042 | 130713015 |
| CRLF2_upstream | X | 1212650 | 1217650 |
| BCL11B_enhancer | 14 | 97382000 | 98876000 |