Lasergene Genomics Suite now includes access to the Variant Annotation Database (VAD) for human sequencing data. I recently spoke with DNASTAR Scientist, Dr. Tim Durfee about the VAD to get a better understanding of how the tool works and how it can help genomics and clinical researchers with their variant analysis.

Can you describe what the Variant Annotation Database is?

The VAD is a database resource that contains information on individual positions and alleles across the human genome. It is currently human genome specific. The major purpose of the VAD is to allow rapid prioritizing and ranking of the large number of variants found in any given sample relative to the reference genome. This can be on the order of thousands of variants for gene panels; tens of thousands for exomes; and millions for whole genomes. This kind of large-scale analysis is critical for the clinical sequencing market.

How can users access the information in the VAD?

Annotation information for each called variant in a specific sample is automatically retrieved from the VAD during project setup in ArrayStar. With the upcoming Lasergene 14.0 release, it will be added to the project directly following assembly and variant calling. The data is accessible in the ArrayStar SNP table and can be used to filter and create gene and SNP sets. For examples on how this can be done, take a look at our tutorial.

What is the source of the annotations in the VAD?

The data is from two major sources: the 1000 Genomes Project and dbNSFP (Database of Human Nonsynonymous SNPs and their Functional Predictions). As the name suggests, the dbNSFP data is on protein encoding positions in the genome. The data are organized into five broad categories:

1. Allele and genotype frequencies from the 1000 Genomes phase 3 data as well as from NHLBI’s Exome Sequencing Project. The 1000 Genomes data is available as global frequencies as well as frequencies for 26 populations grouped into 5 super populations. This data is extremely useful for filtering. For example, if you’re studying a rare disease that only occurs in a small number of individuals, you wouldn’t expect a relevant SNP to occur at high frequency in the population – typically, you filter for variants that occur less than 5% or even less than 1% in the population.

2. Functional impact prediction methods: LRT, MutationTaster, PolyPhen-2 (two models) and SIFT. The four methods use different strategies to predict whether a given non-synonymous change is deleterious to the function of the encoded protein.

3. Evolutionary conservation scoring systems: GERP++, SiPhy, PhyloP and PhastCons. These methods use sequence alignments of the human genome with the corresponding regions of other organisms to produce scores of how conserved each particular base is across evolution. In coding regions, the more evolutionarily conserved the particular base is, the more likely having that base in that position is important for the function of the encoded protein. Some methods (e.g. GERP++) can also be used to assess the importance of bases outside the coding regions.

4. Pathogenicity information from ClinVar: ClinVar is a central repository hosted by NCBI that catalogs and reviews human variation and its connection to disease.  The VAD uses the clinical significance field to allow filtering on different classifications including Benign and Pathogenic.

5. Miscellaneous information: The VAD also contains other types of information such as links to dbSNP Uniprot and Interpro that allow the user to easily retrieve additional data from those resources.

What are the advantages to using the VAD over a user’s own database or VCF file?

If a user has huge VCF files with the annotations, they would have to manually go through each position and retrieve the relevant information for that allele. With the VAD, all the annotations are automatically retrieved and readily available for filtering. The VCF is more useful as a record file of all the variants and their annotations that can be shared between applications.  For example, a VCF of alleles of interest produced by ArrayStar can be used by SeqMan NGen in subsequent assemblies to report on those positions.

How does this compare to other tools on the market today?

The major advantage of DNASTAR’s Variant Annotation Database is the seamless connection with the assembly and variant caller. With open source software, you have to first run the assembly, do the variant calling with a separate program, and then use yet another tool to add the annotation information. There is often a steep learning curve with each of these tools, which can make the overall process laborious. The DNASTAR pipeline integrates all these steps into one suite and allows for multiple sample comparison and filtering. Additionally, we provide the most accurate assembly and variant calling.

Want to learn more? Check out our variant analysis workflow page to see videos and benchmarks on NGS assembly and variant analysis in Lasergene Genomics Suite.

Microscopic view of the Ebola virus

The Ebola epidemic in West Africa has proven to be a devastating reminder that humans as a race are not immortal- we are not exempt from catastrophic biological diseases capable of wiping out entire villages.

In 2014 Kent Brantly and Nancy Writebol were helping patients in Liberia when the two Americans were infected with Ebola. That is where Dr. Michael Pauly and Mapp Biopharmaceutical, Inc. stepped in with ZMapp™. ZMapp™ is an experimental drug for treating Ebola, which at the time, had only been tested on animals. With no better alternative, Brantly and Writebol were given the drug cocktail, comprised of three antibodies. Both patients recovered fully after the drug was given. Today, ZMapp™ is in clinical trials in West Africa.

I sat down with Dr. Michael Pauly to learn more about this experimental drug, his company’s response to the Ebola epidemic, and his experience using DNASTAR software.

Tell us a bit about your recent work with ZMapp™:

A series of events came together, in terms of our development and the current outbreak. We joined forces with a group in Canada screening combinations of antibodies and generated some good animal data.  One of the antibody combinations appeared to be unusually efficacious in animal trialing. It was quite an unusual event that all happened about a year ago, coincident with the increase of this outbreak.

How exactly does ZMapp™ work to fight the virus?

A lot of information is still being found out. The three antibodies in ZMapp™ interact directly with the virus. The degree to which they interact with different epitopes or proteins on the surface of that virus, and exactly how they interact, whether they are directly neutralizing, or whether they enhance the secondary immune response that people have, is still unclear. Something we’ve seen in the primate data, and something so interesting about cocktails, is that by themselves, the individual antibodies don’t show anywhere near the kind of effectiveness that they do in combination. The whole idea of antibody cocktails has really been bolstered as a part of this result.

What was it like to be involved with the Ebola crisis? It’s been very dramatic, as you said, and pretty devastating.

Ultimately for most everybody, it’s just tremendously motivating. Nobody has any trouble working all day and all night when faced with those sorts of issues. You’re just trying to make an effective product and that part makes your work life and the rest of your life pretty simple. Your work becomes very personal and in that way, quite meaningful. It gives everybody an extra added boost to work hard and make something that works.

What are some of the unique challenges you face in your work?

We are always trying to relate antibody structure to function, and we spend a lot of time trying to figure out what makes an antibody express well or perhaps not so well. Another set of challenges is related to speed and throughput. We’re screening a lot of antibodies and reagents, and anything we can do to increase that throughput and, in the case of DNASTAR software, anything we can do to more effectively model and screen a sequence in order to understand structure function is helpful.

How does DNASTAR software help you with those challenges?

We use the DNA side of your software all the time. We use it daily for DNA analysis. We have our own particular database of all the sequences we’ve made, and almost all the protein sequence that we currently use is in DNASTAR- based format.

For a long time scientists just made antibodies in cell lines without knowing what the sequences were; they just purified the antibodies and used them as reagents. There are still cell lines going back ten years or more that are making good antibodies, but people don’t know the sequences of them. We sequence cell lines that express antibodies we don’t know the sequence of, so we use degenerate primer pools to try to retrieve the antibodies that are being expressed by the cell line. We use SeqMan Pro to assemble traces, make contigs, and try to find what look to be antibody sequences that are being expressed by various cell lines.

Another thing we have always appreciated is DNASTAR has good support. You can call somebody and get help easily.

Would you like to see your organization featured on the DNASTAR Blog? Contact thomase@dnastar.com to schedule an interview. We love learning about the important work of our customers!

Most variant callers provide probability scores for each detected variant. However, these values alone do not tell you how reliable the overall results are. Just because a difference from the reference sequence is statistically significant doesn’t necessarily mean that it is an accurate call.

How can you know if your results are accurate?

In most studies, especially when looking for rare mutations, having a reliable reference set with known variations isn’t feasible. To test the accuracy of NGS alignment and variant calling in Lasergene Genomics Suite, we used SeqMan NGen to align whole human exome data from the Genome in a Bottle Consortium (GIAB) to the human genome. Because this is a well curated data set, we were able to compare the variant calls to the “answer” provided by GIAB. We also performed alignment and variant calling in several other software packages using the same data and comparable settings. We then looked at three metrics:

1. Sensitivity – This is also known as the true positive rate, and is the ratio of correctly identified variants to the total known variants in the reference set. The higher the sensitivity, the greater the likelihood that a variant in the sample will be identified by the software.
2. Specificity – Also known as the true negative rate, this is the ratio of non-variant calls to the total number of positions in the reference set that are known to be homozygous with the reference sequence. Specificity is inversely related to the number of false positives.
3. False Discovery Rate (FDR) – This is the ratio of false positives to all variant calls made by the software. The FDR value for a variant caller allows you to understand how many variants in your project are likely to be false positives.

Because an accurate alignment is a necessary precursor to accurate variant detection, these metrics also help you understand the alignment accuracy from various software pipelines.

Here’s a peek at the results we obtained:

Check out our new Accuracy page to see the full results of these comparisons across numerous exomes and other sample data sets and learn how Lasergene Genomics Suite stacks up against the competition!

Today, researchers often find they must wait for the limited availability of biostatisticians and bioinformatics teams. However, with the right tools, it is hoped that the solutions to Mendelian and complex diseases will be discovered by investigators at their own desktop computers.

DNASTAR combines the computational power required to prioritize relevant factors and visualize correlations in an easy-to-use, integrated software pipeline that puts the power of association studies into the hands of clinical researchers.

A Powerful Integrated Software Pipeline

The number and size of NGS data sets that are needed to conduct association studies can pose some challenges. First, the massive amount of raw data (typically 10-30 GB for a single exome and 300-400 GB for a whole genome) requires substantial computer resources for processing as well as for storage and management. Second, there is a series of computational tools required:

1. A large capacity NGS reference-guided assembler
2. A variation detection module
3. A variant annotation module
4. A visualization package for inspecting alignments and variant calls
5. An analytics module for comparing variants across samples including statistical analyses and discrete filtering

Stringing together and running the software tools needed to accomplish these tasks can be quite a hurdle. But, with the integrated Lasergene Genomics Suite, the flow of data is facilitated with an easy-to-use, intuitive, graphical interface. The suite consists of three programs: SeqMan NGen, SeqMan Pro, and ArrayStar.

Case Study:  Kabuki Syndrome

As a demonstration of DNASTAR’s pipeline, a rare Mendelian disorder known as Kabuki syndrome was used. Exome data sets were obtained through dbGaP. These data sets were from the published Kabuki syndrome study (Ng et. al. Exome sequencing identifies MLL2 mutations as a cause of Kabuki syndrome. Nat. Genet. 42, 30-35 (2010)). The syndrome, which is caused by autosomal dominant mutations, is rare with approximately 400 cases reported worldwide.

Ten case and eight control exome data sets were independently aligned to the human genome reference sequence using SeqMan NGen which also identified and annotated variants. Variants from each assembly were then loaded together into ArrayStar resulting in over 5.7 million independent positions located in about 32,000 genes across all samples after coalescing. The samples were then organized into two groups, Kabuki and Control, to facilitate subsequent filtering.

We first filtered at the variant level making three assumptions based on knowledge of the disease: 1) causal mutations would be non-synonymous changes, 2) causal mutations arose de novo so variants would occur in only one case sample and 3) no control sample would have any of the mutations. Stringent quality metric thresholds were also imposed to reduce noise. Over 11,000 variants in 6,352 genes met the criteria and were saved as a “SNP set.”

This SNP set was then used as the variant pool in a second filtering step. This time to identify genes with mutations that met the following criteria: 1) mutations were inactivating (nonsense or frameshift), 2) they were rare and therefore not in dbSNP and 3) they were dominant and therefore occurred as heterozygotes. 845 genes met those criteria in at least one case sample. However, by increasing the level of detectance to 7 of 10 case samples the number of candidates was reduced to one, MLL2, consistent with the results of Ng et. al.

The DNASTAR software makes this type of filtering easy for researchers with an intuitive filtering interface:

Conclusions

Exome sequencing has revolutionized our ability to detect common, rare and private variants in the coding genes of an individual. By sequencing case and control cohorts and then comparing across the spectrum of variants, the genetic causes of Mendelian and complex diseases are being uncovered. NGS technologies and facile software pipelines that integrate assembly, variant calling/annotation and association analyses are essential partners in this endeavor.

Tell us about your work!

My research has primarily focused on the molecular and cellular biology mechanisms of rare pediatric diseases, neuroinflammatory pain and neuromuscular development. Quantifying mRNA expression levels across a broad range of genes has been a critical component within my projects. I use the Lasergene software to optimize everything from primer selection to analysis of our sequence data. The biggest challenges are linking the species-specific genetic and protein information to various pathways.

How has DNASTAR software helped you?

The Lasergene software has really helped us with our bioinformatic-driven projects. We are constantly exploring both novel drug targets and working to gain a better understanding of current targets of interest. Particularly when working on physiological and pharmacology experiments, we need to have a high degree of confidence in our molecular work. The Lasergene software gives that to us, especially with gene sequence and protein analysis.

What does DNASTAR software do best, in your opinion?

Data analysis, especially with sequence alignment and working with results from the sequencer. I have also come to appreciate the flexibility of the software with primer design. I have also recently started working with Protean 3D and find it to be extremely helpful with understanding poorly published proteins of interest.

Can you speak to DNASTAR’s support for you and your work?

While at the Medical College of Wisconsin I managed a large departmental network because of my familiarity with the program. While things typically ran smoothly, any questions I had were answered same-day, which made my work incredibly easy.

The resources provided by the National Institute of Standards and Technology (NIST) Genome in a Bottle (GIAB) consortium promise to greatly improve the reliability of genetic assays. With these tools, laboratories can integrate performance measures directly within the workflow of their testing operations.

Nephropathology Associates, Inc. (Nephropath™), a leading U.S. laboratory in the interpretation of kidney biopsies, was motivated to use the NIST materials by the need to demonstrate proficiency in their NGS platform for purposes of CAP/CLIA certification. They were encouraged to look into GIAB by a representative from Illumina and, after a discussion with Justin Zook at the 2013 ASHG conference, decided that using the NIST data was the best option for them. The approach was also appealing because it would provide a measure of the lab’s accuracy as they would be able to compare their data with that of others who use the same controls.

As part of a collaborative project between Nephropath and DNASTAR, a new workflow has been added to DNASTAR’s assembly and variant calling software that supports use of the GIAB call sets.

The workflow is designed to work with a “gold standard” control of the user’s choice, such as the set of reference materials for the HapMap/1000 Genomes CEU female NA12878 developed by the GIAB consortium, as shown in Figure 1.

Figure 1. DNASTAR’s integrated Validated SNP Caller workflow used with NIST GIAB gold standard reference materials.

The purpose is to validate the efficacy of a procedure from sample prep through sequence analysis. At the end of the workflow, the lab obtains an automatically generated statistical report detailing the assembly sensitivity, specificity, and accuracy calculated according the ratios described in Table 1.

Table 1. The Validation Report calculations.

Nephropath is currently using the Illumina MiSeq and Agilent SureSelectQXT with custom probes for 301 genes involved in kidney disease. They use DNA from NA12878 purchased from Coriell Institute as a sequencing control on every run. Each run is a pool of 9 samples plus the control sequenced with the paired-end MiSeq® Reagent Kit v3 (150 cycle). The NA12878 control FASTQ files generated after the run are loaded into DNASTAR’s SeqMan NGen® software for mapping/alignment against the human genome reference sequence and variant calling using the “Templated assemblies with control” option. To delimit the regions of the genome used for validation, Nephropath uses a BED file of either their entire targeted region or one containing an intersection between the GIAB high quality regions and the targeted regions. The latter is preferred when the most accurate statistics are required as suggested in this README file. In this way, the NA12878 variant call set VCF file gets subsetted down to just the targeted regions using whichever BED file is selected. After the assembly is complete, every position specified by the BED file, including both variant and reference calls, is checked against the subsetted control VCF to determine true/false positives/negatives. Based on these annotated variant and reference call sets a validation report is generated by the DNASTAR ArrayStar® application, providing various statistics achieved at different sequencing depths and probability thresholds. An excerpt of such a report is given in Figure 2.

Figure 2. Nephropathology Associate’s Kidney Disease Gene Panel: Excerpts from a NA12878 Validation Report. Data provided by Marjorie Beggs (Nephropathology Associates) includes 301 genes from 13 renal disease categories.

The pipeline, along with early results, were presented at the recent GIAB consortium workshop in a roundup of emblematic case studies on using the GIAB materials.

Nephropath, in collaboration with DNASTAR, was recently awarded an SBIR phase I grant to further develop this workflow and software for clinical use. The long-term goal is to implement a fast, accurate and integrated workflow for clinical NGS.

For more information on the new validation workflow, download , “SeqMan NGen is a High Accuracy NGS Assembler: Assessment with NA12878 Reference Materials,” from the DNASTAR website.

Tell us about your work!

Currently, we do a lot work in Africa with specific focus on Mycobacterium tuberculosis (MTB) detection and next-generation sequencing (NGS). We develop molecular reagents and products that simplify and enhance molecular detection from the point of sample collection, detection, and DNA sequencing. We also do work with influenza viruses and have recently done some whole-genome MRSA sequencing as well.

How has DNASTAR software helped you with your research goals?

In 1999, I established the U.S. Air Force’s molecular influenza strain surveillance program and was first introduced to DNASTAR by the CDC. Since then, I’ve never let go. I prefer DNASTAR over other bioinformatics software because it’s so simple to use. I have observed and participated in the evolution and maturation of DNASTAR software from simple gene alignments of DNA off automated slab gel sequencers to today’s cutting edge, multimillion read NGS assemblies! DNASTAR continues to mature alongside rapid advances in sequencing technologies.

What does DNASTAR software do best, in your opinion?

Since we do a lot of barcoding/indexing of patient isolate samples, I like the ease with which I can perform a multi-patient analysis within a single SeqMan NGen assembly-the software enables you to quickly switch within barcoded/indexed patients. I also like that with SeqMan NGen I can quickly input a reference gene or gene panel of interest against a full genomic library file to quickly assess for mutations in genes of interest. Sometimes you just want to look at specific genes and you don’t always need/want the hassle of dealing with an entire 4.2 million bp MTB genome!

Can you speak to DNASTAR’s support for you and your work?

The DNASTAR team works closely with Life Technologies and Illumina to simplify the analysis of NGS data for the everyday scientist. Most scientists are good at empirical experimentation and molecular biology testing but a bit intimidated by the size of raw data output and the bioinformatics aspects of NGS. DNASTAR bridges the gap between 2-3 gigabytes of raw NGS sequence data and a thorough genetic analysis containing graphical alignments, mutational reports, and phylogenies.   For example, from incredibly large raw MiSeq data files containing 24 indexed MTB genomes I can easily use the tools in DNASTAR to assemble contigs and analyze TB isolates for mutations in antibiotic resistance genes. From here it’s simple to generate SNP reports, multiple sequence alignments, and phylogenetic trees.

Is there anything else you’d like to share with us?

When I get into trouble I find it easy to contact DNASTAR technical support. Matthew Keyser and his team are always available for troubleshooting and resolving user issues.