Bio-IT World’s Editors’ Choice Award: XAbTracker and SeqAgent

By Rory McCann

June 24, 2016 | Each year, the Bio-IT World Conference & Expo welcomes innovators from around the world who come to share their best practices, and glean insight from others. It is an event teeming with challenges, solutions, and countless opportunities to look at discoveries in the bio-tech industry in new ways. The Bio-IT World Best Practices Awards recognize and showcase the teams that are not only recognizing areas in need of improvement, but finding applicable solutions as well. This year, XOMA’s applications XAbTracker and SeqAgent stood out for doing just that. Their team targeted a bottleneck in the antibody discovery process and developed two software programs specifically designed to improve data analysis with both speed and efficiency.

XAbTracker and SeqAgent took home the Bio-IT World Editors’ Choice Award for providing an alternative method to what Mark Evans, Associate Director of Technology Innovation and Bioinformatics, calls an antiquated and inefficient system of analysis. Despite the strides made in antibody discovery via phage display, Evans says DNA sequence analysis and candidate selection can still eat away at valuable time. His answer to this is the XAbTracker and SeqAgent. These sequence analysis tools streamline the antibody discovery process by organizing the data for review and by identifying structural features and producing annotated alignments, while also providing a flexible workflow and efficient data management. XOMA’s team believes its product will save both precious time and resources.

SeqAgent is a standalone sequence analysis program. The user uploads data in the form of hundreds of ABI files to a server, where SeqAgent analyzes the files and tests them for their sequence quality. It identifies similar and unique sequences, and also labels, organizes, and analyzes for patterns. The sequences are clustered according to the Levenshtein algorithm, which measures the data’s similarities. The users are able to view these as a visual summary that can be changed at the user’s discretion to optimize their selected analysis, whether it is accessing individual sequences or looking at sequencing patterns as a whole.

XAbTracker is responsible for managing data. It can receive an enormous data input, and then organizes the information by monitoring data for expected patterns, as well as new trends and abnormalities. It has several visualization options to view data and conduct analysis. It then gives the user accurate feedback in real-time, summarizing results for each assay and providing fast and reliable information, nearly eliminating the need for further sample analysis. This process can streamline analysis by sifting through routine data that yield no impressive outcomes and highlighting new discoveries that propel the entire process forward. Together, these programs can reproduce what used to take hours of analysis and yield a greater amount of useful data at a lower cost and higher quality.

In addition to their innovative processes, the programs stand out for having flexible interfaces and being cost-effective. The in-house developers designed them to be web-based and open-source, making both XAbTracker and SeqAgent user-friendly and accessible on various devices including PC, tablet, and smart phone. XOMA believes the applications also cut costs and increase the number of samples for analysis and the yield of useful sequences.

XAbTracker and SeqAgent won the Editors’ Choice Awards for their improvement for workflow and data analysis, but according to Evans, XOMA isn’t not done yet.

Since submitting to Bio-IT’s Best Practices competition, his team has added a new feature to the XAbTracker application that enables users to look at the cumulative results in a new way. “We had that before,” Evans said of the review. “But it was a very ugly view.” It worked, but it was not efficient. The redesigned feature allows researchers to view their samples in a new Venn diagram system, but according to Evans, it’s more than just two overlapping circles. The Venn diagram consists of multiple circles; each overlap is interactive and can contain several samples. It is a responsive filter that takes the data, a list of samples, and converts what would normally be listed as hundreds of values in a table, to data values in a visually dynamic view that is both efficient and responsive. The redesigned feature will improve the process with cumulative results that are straightforward and easily accessible.

Evans says, “It’s very nice and it’s a huge improvement and aesthetically it’s much preferred, and we just rolled that out a few weeks ago, and it’s been very well received.”

Evans says they are now moving towards mining information from the sequencing data they’ve been collecting. The goal is to apply more bioinformatics to build predictive models from the sequences they already have. However, Evans says, “The data structure we have is not optimum for data mining.” XOMA is addressing this issue and hopes to improve protein engineering software by creating a new class of software to support both protein engineering and antibody bioinformatics processes. Evans and the team at XOMA are hopeful these new advancements will hit the bio-tech community strong and soon, but for next year’s Best Practices Awards? We’ll have to wait and see.