The software comes with 2 installers: (1) SureCall installer (2) GenAligners (contains BWA, BWA-MEM).

Transcription

1 Release Notes Agilent SureCall 3.5 Product Number G4980AA SureCall Client 6-month named license supports installation of one client and server (to host the SureCall database) on one machine. For additional client only installations that connect to the same database on the central server, additional copies of this license are needed. There is no limit on the number of free-of-charge licenses provided. The software comes with 2 installers: (1) SureCall installer (2) GenAligners (contains BWA, BWA-MEM). Overview SureCall is a desktop application combining algorithms for end-to-end NGS data analysis from alignment to categorization of mutations. SureCall addresses the critical need for an easy-to-use analysis tool that incorporates the most widely accepted open source libraries and algorithms, augments them with tools specific to Agilent assays and deploys them in a convenient and user-friendly manner. SureCall provides four different types of analysis: Single, Pair, Trio, and OneSeq CNV and Mutation analysis. Analysis in SureCall begins with raw reads from Illumina HiSeq/MiSeq or aligned reads from Ion Torrent sequencing of genomic DNA enriched with HaloPlex or SureSelect. After removal of the adapter sequences, the reads are aligned to the genome (BWA-MEM, BWA). Single Sample analysis Run a single sample analysis when you want to find mutations and insertions or deletions (indels) in individual samples. SAMTools or SNPPET, an in-house algorithm developed specifically for the detection of low allele frequency variants, can be used to call mutations. For samples that were target-enriched using Agilent's HaloPlex HS, duplicate reads will be flagged, allowing for an even more accurate detection of alleles at low frequencies. Several tools are then used to provide input for the mutation classification. Each mutation is evaluated based on its location, amino acid change, and effect on protein function (SIFT). Further information regarding the mutation is then aggregated from various public sources, including NCBI, COSMIC (Catalog of Somatic Mutations in Cancer), PubMed, and Locus-Specific Databases. In addition, SureCall also supports variant annotation with NCBI ClinVar files with a local database source. After collecting the various inputs for classification, the proprietary mutation classifier evaluates the significance of the mutation following default or customized guidelines. Each mutation is then categorized with the user triaging each mutation and reviewing supporting evidence in the built-in viewer, including raw data and confidence measures, as well as links to external databases such as OMIM and dbvar, dbsnp, etc. Pair analysis Select pair analysis to determine copy number changes in a sample. You will need to use a sample that does not have a copy number change in your region of interest as a reference sample. You can also use

2 pair analysis to find mutations and insertions or deletions (indels) in tumor versus normal pairs to determine somatic mutations in the tumor sample. Trio analysis Select trio analysis to find mutations and indels in a trio of samples, typically mother, father and child. The analysis focuses on de novo mutations, i.e. mutations that are only found in the child, and mutations that are homozygous in the child but not in either parent. OneSeq CNV and Mutation analysis A OneSeq analysis simultaneously finds CNVs, copy-neutral LOH, point mutations, and indels in a single sample. The OneSeq workflow type is only suitable for samples that were target-enriched using Agilent's OneSeq kits. Copy number changes are detected by comparing an experimental sample to a known reference sample. The in-house developed SNP calling algorithm SNPPET is used to call point mutations and indels. The high minor allele frequency SNPs covered by the OneSeq backbone design are used to determine copy-neutral LOH. New Key Features of SureCall 3.5 Tools for adding calls to a sample: Starting in SureCall 3.5, the Triage View has tools that allow you to add a SNP, deletion, or insertion call to a sample. For samples analyzed in a OneSeq analysis, you can also add CNV and LOH calls. Tools for editing CNV and LOH calls: For samples analyzed in a OneSeq analysis, the SureCall 3.5 Triage View allows you to edit the start and stop positions assigned to gains, losses, and LOH intervals. For gains and losses, the copy number assigned to the call is also editable. Rule sets for automatic call management: SureCall 3.5 enables users to design rule sets, which automatically suppress or re-categorize specific mutations identified in a sample. Each rule set is comprised of one or more rules, and each rule defines a mutation and a treatment for that mutation (suppress or re-categorization). Rule sets are created from the Supporting Files screen or from Triage View, and can be applied to a sample from within Triage View. Auto-suppression based on tracks: In SureCall 3.5, the tools for suppressing mutations in a sample have been expanded to include an Auto Suppress feature that suppresses mutations that overlap an interval in a user-selected track. When applying auto suppression in a OneSeq sample, the degree of overlap between the mutation interval and the track interval is customizable. Enhanced SNPPET caller: The SNP caller used in SureCall 3.5 has an improved ability to detect both homozygous and heterozygous low-frequency SNPs and improved overall sensitivity and specificity. Ability to import multiple FASTQ files per sample: Starting in SureCall 3.5, when setting up an analysis workflow, you can import multiple file pairs for a single sample, which is especially helpful for samples that were loaded on multiple lanes of the sequencer. Support for Windows 10: SureCall 3.5 is compatible with the Windows 10 operating system. User-controlled management of memory allocation: Administrators of SureCall 3.5 can manage the amount of memory on the machine that is allocated to the SureCall application and to running SureCall workflows. System Requirements See the SureCall installation guide (P/N G ) or the SureCall web site for minimum and recommended system requirements and installation instructions.

3 Installation Instructions New installation Refer to instructions in Points to note for upgrade from older version of SureCall to SureCall 3.5: Upgrade is only supported from SureCall 2.1 or SureCall 3.0 released version (v or v or v3.0.3) to SureCall 3.5. SureCall versions prior to 2.1 (i.e. 1.0, 1.1 and 2.0) will not be upgraded to 3.5. The user must first upgrade from version 1.0/1.1/2.0 to version 2.1 version using SureCall 2.1 installer. PostgreSQL server of earlier version will be upgraded to 3.5. There will not be a new server installation. Hence, after upgrade, server and client installation folders will be saved to different locations on disk. If an uninstallation is required, the client and server will need to be uninstalled separately. Upgrade instructions from earlier (v2.1 or later) SureCall version 1. Uninstall GenAligners v1.1 or v2.0 and install new v3.0 version of GenAligners. 2. Double-click the Agilent SureCall x.exe file to start the installation wizard. You will be prompted that a version of SureCall client already exists on the local machine. 3. Click OK to proceed with uninstalling the existing Agilent SureCall client. The Uninstaller of the existing SureCall installation is launched. 4. Click Next to proceed. 5. Select Uninstall specific features, and click Next. 6. In the top panel, check the Client checkbox, and click Uninstall 7. After client uninstallation of earlier version is complete, v3.5 installation will be resumed. 8. Select Both Client and Server option and click Next. The installer will prompt that SureCall server already exists and it will be upgraded to latest version. 9. Click OK and proceed with the installation. 10. The installer will install SureCall 3.5 client application and upgrade existing SureCall server to 3.5. Default Analysis Method changes Adapter Trimming is performed in all default SureSelect methods. User can choose Enrichment Method while importing Fastq files and appropriate adapters will be trimmed. With this change separate default QXT analysis methods are no longer needed and have been removed. The 'Disable Adapter Trimming for SureSelect' parameter, which was used to turn off trimming for SureSelect XT in earlier versions, is thus no longer applicable and has been removed as well. Unused 'Fix Mate' parameter deleted from post-alignment option. Bundled annotations have been updated and new version is marked V3 Annotation version. This new version 'V3' is the default for all default analysis methods. You can revert to an older version by choosing it in a custom analysis method. Following parameters removed from all analysis methods: o Contamination Fraction o Ploidy o No minimum mapping quality for read o Read Depth filter o Variant Quality filter o Strand ratio filter

4 o Low frequency specific parameters: Quality value threshold (low frequency) Min-AF(low-frequency-Homozygous) Min-AF(low-frequency-Heterozygous) Min mapping quality for read (low-freq) Min quality for base (low-freq) 'Minimum Allele frequency' and Minimum number of reads supporting variant allele parameters added in all analysis methods. 'Amplification level per barcode' parameter added in all HaloPlex HS analysis methods. Following parameters removed from all SureSelect analysis methods: o Minimum average distance of variant from 5 end o Minimum average distance of variant from 3 end For Pair CNV analysis methods, the parameters Min adaptive interval size and Max adaptive interval size replaced with the parameter Adaptive interval size for CNV calls, and the default value for Minimum Read depth for Reference parameter updated to 10. Parameter names for SOMATIC/Denovo variants changed. Updated default value for a read depth parameter supporting SOMATIC/Denovo to 10. For OneSeq analysis method, Split aberrations across p-q arms parameter added. Default value of Minimum average distance of variant from 5 end parameter updated to 5. Filter indel and corresponding parameter removed. Default value of Variant call quality threshold parameter is now 100 instead of 5. HaloPlex and SureSelect analysis method parameters are separated. Default OneSeq Full Design CN analysis method removed. The variants are called using the full design and backbone probes are used for copy number calculation. SureCall 3.5 result differences Variant results in SureCall 3.5 are expected to be different as compared with earlier SureCall release build (v3.0) because of: 1. Fixes to issues in variant detection 2. Changes in analysis method parameters 3. Improvements/changes in track handling as per BEDOPS (see changes affect known variant results, track filtering, and track annotation 4. Use of LocateIt for default masking/trimming of HaloPlex data 5. Changes in pre-processing and molecular barcode handling step, which results in QC metrics changes 6. Deprecation of the BAQ flow, which is not updated for newer parameters Issues Fixed in SureCall In some samples, when multiple variants are identified at the same position, only one variant is reported in the results. (TT#228549) 2. In Triage View for Trio analysis, the gene name is not displayed for some variants. (TT#235947) 3. In the OneSeq Triage View, data plots overlap the log ratio scale or B-allele frequency label. (TT#238427)

5 4. On some machines, common storage folder is not shared after SureCall installation. (TT#241134) 5. In Triage View, if users search for aberrations in other samples, open one of the samples, then search again for a different aberration and open the same sample again, the aberrations from the first search are still highlighted in that sample. (TT#243099) 6. In Triage View, the HGVS values are swapped for some variants. (TT#243518) 7. Repeatedly opening the OneSeq Triage View more than 5 6 times causes the application to become unresponsive. (TT#243520) 8. Custom attribute section and free comment section are not visible in the custom NGS report. (TT#243988) 9. In OneSeq Triage View, no log ratios or aberrations are displayed in the psuedoautosomal regions. (TT#244062) 10. In OneSeq Triage View, sometimes the link to OMIM website opens an error message. (TT#244258) 11. In the OneSeq Triage View, the Reset All option does not work properly. (TT#244545) 12. In Triage View, for a sample analyzed with a known variants file, when users move a variant designated as NOT_CALLED_OUTSIDE_REGION then check in the sample and close Triage View, the Triage View cannot be re-opened. (TT#244560) 13. In OneSeq analyses, sometimes CNV aberrations are called that are smaller than the minimum size specified in the analysis method. (TT#244652) 14. In particular cases, when users import a new HaloPlex design, SureCall erroneously defines it as a HaloPlexHS design. (TT#245886) 15. For samples containing a 3-base deletion, the reference allele is reported as if it is a 5-base deletion and, consequently, the mutation effect is incorrectly designated as a frameshift. (TT#245953) 16. In OneSeq Triage View, the tools for creating and using bookmarks do not work properly. (TT#245966) 17. In Triage View for a HaloPlex sample analyzed in a Pair CNV analysis, not all columns display correctly in the Mutation table. (TT#246075) 18. In OneSeq Triage View, the hashed line option in the Preferences dialog does not work. (TT#246271) 19. In OneSeq Triage View, settings for tracks as determined in the Preferences dialog are not in sync with the track display. (TT#246272) 20. In OneSeq Triage View, some of the color settings options for tracks do not work as expected. (TT#246273) 21. In Triage View, when users search for aberrations in other samples, if they then try to load the results for those samples that contain a match, they receive an error message if those samples are from an externally loaded VCF file. (TT#246290) 22. In custom NGS report, hidden attributes are displayed as blanks. (TT#246490) 23. In Pair analyses, the reported forward reference allele count is incorrect. (TT#246759) 24. When creating a reference file from the Supporting Files screen, import of the reference sample file fails if application is closed during import. (TT#246827) 25. Downloading a new HaloPlex or SureSelect design fails if annotations are not successfully downloaded. (TT#247110) 26. In the Multi-sample window for a Trio analysis, when exporting table data, the tab names in the resulting file are incorrect. (TT#247111) 27. In Triage View, for a sample analyzed with a known variants file, the Type and Comments column are not populated on the Known Variants tab. (TT#248278) 28. The results for different samples that contain the same indel show different references at the indel position. (TT#248463) 29. The analysis method parameters for detecting low-frequency variants do not filter the results as intended. (TT#248572) 30. In Triage View, in some cases in which the sample contains a large number of mutations, not all mutations that are displayed in the genome viewer are not listed in the Mutation table(tt#249816) 31. In the QC report for samples analyzed from single-end reads, the number of duplicates is erroneously reported as 0. (TT#259108) 32. In the Triage View for a Pair CNV analysis, if a user re-orders the columns on the CNV tab, the data in the columns does not get properly re-ordered. (TT#259965)

6 Known/Open issues 1. Searching the Mutation table is slow for samples with a large number of variants. (TT#218779) 2. In OneSeq analyses, sometimes the call is extended by 1 probe beyond what was visually selected as the boundaries of an aberration. (TT# and ) 3. In Triage View, the regions below threshold do not include the last base of analyzable target region. (TT #248406) 4. Ensembl IDs have been updated in Ensembl database for new genome builds, etc., and some of the older IDs have been retired. For these retired IDs, the hyperlinks in the Mutation table points to a blank page. (TT#252572) 5. In a custom analysis method, setting the quality value cutoff to zero in the trimming parameters turns trimming off. With this setting, no trimming is performed and QC metrics corresponding to trimming are not available. However, Agilent recommends that users do perform quality trimming for low quality bases at the ends of reads. (TT#257037) 6. In certain OneSeq analyses, the CNV table in the Triage View is reporting data points not displayed in the genome viewer. (TT#257822) 7. In analysis method for Trio analysis, the parameter "Min number of reads supporting DeNovo variant" is applied to total read depth and not to reads with variant allele. (TT#258824) 8. For some Ion Torrent samples, the same deletion is reported multiple times in the Mutation table. (TT#259319) 9. The summary information for various effects is provided as a guideline in the job summary and includes count of variants with different annotations. However, the count for variants of one particular functional class is less than the count shown in the Mutation table. The Mutation table contains the correct count for these variants. The other counts in the job summary are correct. (TT#258994) 10. In OneSeq Triage View, some of the SNPs do not have Design Type annotation, i.e., Backbone, Targeted, or Overlap. (TT#260476) PR