Mojca Strazisar on the Challenges in Sample Prep Workflows

Mojca Strazisae Photo credit: ©VIB Ine Dehandschutter

Mojca Strazisar is an expert scientist at the Genomic Service Facility in the VIB-UAntwerp Center for Molecular Neurology, with a focus on genomics technologies. She moved to Belgium six years ago to start working as a postdoc in the department. Roughly two years ago, she took over the role of head of the Genomic Service Facility, a position she very much enjoys.

Q: What does your laboratory do?

A: The Genomic Service Facility (GSF) is one of the core facilities affiliated with the VIB (Flemish Institute for Biotechnology) and the University of Antwerp. We provide Sanger sequencing services to academia, to institutes, and to third-party clients, and since 2013, we have also been providing next-generation sequencing services (NGS) on Illumina platforms. We are based in Antwerp in the lovely green surroundings of the Campus Drie Eiken.

Q: What key research projects are you supporting at the moment?

A: As a service facility, mainly providing Sanger sequencing services, sometimes we have no idea about the relevance of the projects. We receive orders and samples, which we process, and deliver the results. But we do stay connected with the science, as we are located within a research environment. Most of our NGS service helps researchers unravel interesting and important mutations in a human genome that are causing, contributing to, or associated with different types of neurological disorders. (More information can be found here, http://www.molgen.vib-ua.be/Public/About/Default.cfm.)

Q: How many staff members do you have?

A: Now we have nine people on staff, but that always shifts because we can house students and interns, so sometimes the exact number of GSF-affiliated members changes. Q: What main technologies do you use? A: For the Sanger services, we use three sequencers (ABI 3730xl) with very high throughput. Because we had an overcapacity, we started to provide Sanger sequencing as a service. First we started off through [word of mouth] to other colleagues, institutes, and departments, and in due time we had set up a sequencing facility with automation and advanced technology to be able to provide NGS services. There, we mostly perform targeted resequencing, exome sequencing, and also RNA sequencing.

One of the most crucial parts of our work requires thorough quality control in many steps of the NGS protocols. When a sample is received, we check the quality using UV-Vis and fluorescence spectrometry, but that provides insufficient information on the sample status; therefore, we use fragment analyzers, such as Advanced Analytical Technologies’ (AATI) Fragment Analyzer, to quantify and evaluate samples, libraries, and the success of the protocols. Knowing the integrity and size of the fragments, whether they’re original sample or library, is the crucial step in all NGS-related protocols. As we are working on mid-highthroughput service, the Fragment Analyzer gives us the most flexibility in terms of type of analyses and number of samples.

Q: What kind of sample prep is involved in the work you do?

A: Besides Sanger sequencing, which is fully automated with the exception of a few steps, we also perform DNA and RNA extractions using PSS systems. In the NGS pipeline, we do everything more or less by hand or half automated. For example, PCR purification, aliquoting, and pooling are all done on [Beckman Coulter] Biomek instruments (NXp, FXp, and NX), but library prep is done by hand. For the NGS service, specifically exome and RNA sequencing, researchers or clients provide samples and request a service, and we proceed from there. We perform all the quality control needed; of the sample, library, and sequencing run, and we construct the libraries and sequence. We also have an in-house bioinformatics unit that analyzes the data.

Q: How many samples do you deal with in a month, on average?

A: That is difficult to say, as some of our service is project dependent, but on average we process approximately 50,000 samples per month for the Sanger sequencing and around 3,000 samples per month for the NGS: being targeted resequencing, exome, and RNA sequencing.

Q: What types of samples do you deal with most often?

A: Most of our samples (DNA or RNA) are of human origin, but we can and do process any type of sample—bacterial, plant, or animal. As a service, as long as it is DNA, RNA, or derivative—cDNA, amplicon, plasmid, library—we can process it.

Q: What are the main challenges you face in your sample prep workflows?

A: My biggest pet peeve, if I may call it that, is that sometimes the reproducibility in the different sample prep protocols is low. I think the advancement is not at the point yet to completely understand where this low reproducibility is originating. Also, we try to have zero or close to zero tolerance for errors, so we try to figure out how to unify different sample types—and by different sample types, I mean different samples, different origins, and different extractions—in order to have the reproducible data quality.

What I also dislike at this moment is that there are still some bottlenecks, certainly in the QC [quality control] steps. We try to do the best we can, so we’ve introduced several additional quality controls during the library prep, and those usually are the bottlenecks of the whole protocol. We analyze every sample after every two or three steps of the protocol, and that takes time. It does help to have and use state-of-the-art quality control and sample-handling equipment in-house and working under GLP (good laboratory practices) as well as using a LIMS (laboratory information management system) to diminish problems in reproducibility, failure, and in human errors, but for now those are the challenges we confront on a daily basis.

Q: How do you deal with those challenges?

A: The easiest way is to follow up on what is new on the market. We have to adapt existing protocols based on the origin of the samples or their quality. It is important to understand the origin of the samples, solvents they are diluted or stored in, and the correlation between good results and sample quality. If you start with the sample, not knowing the quality of the sample or the sample prep, for example, and then you end up trying to interpret the sequencing data, you can’t always conclude whether what you observe is of some biological relevance or is the result hindered by a technical problem that occurred somewhere during the sample manipulation.

Q: What major changes have you experienced in regard to your sample prep workflows over the past few years?

A: Monthly, there are novel protocols and novel products on the market. That is actually very, very good, as there is more to choose from. We see better equipment; I’d have to say the prices of the equipment are also getting lower, and more protocols can be run on the same platforms. The protocols and the instruments are updated and adapted to different requirements of different researchers. It seems there are a lot of manufacturers who are really listening to the questions or requirements of the researchers.

Q: What have those changes meant for your lab?

A: Basically, we need less time to do more. The instruments, such as the DropSense 16, DropSense 96 [Trinean], Biomek liquid handlers, and the Fragment Analyzer help us tremendously in our daily work tasks, as they can be adapted to different sample and project requirements and different protocols and use consumables in a non-wasteful manner. We better understand what is happening to the sample, how the results are obtained, and what the results mean. We can follow the quality of the sequencing and the quality of the sample prep more closely. It gives us the possibility to do better and to do more for the same cost.

Q: What are the plans for your lab? How do you expect your sample prep workflow to change going forward?

A: To expand our portfolio of what we are servicing on the NGS side. We want to also get into the sequencing of longer fragments. In practice, we would like to adapt even shorter protocols, better protocols with better reproducibility, so the quality of the data can be more unified. What we really need specifically in our lab, because we are still very connected to the researchers and specific projects, are instruments, such as the AATI’s Fragment Analyzer-12 cap and Fragment Analyzer-96 cap instruments, that can perform mid- to-high-throughput quality control and sample handling. If we have to process a limited number of samples—not one, but also not hundreds at a time—we do not want to waste consumables or time.

Q: What do you enjoy most about your work? Why?

A: For sure solving, I don’t like to say problems, but solving challenges. Figuring out how the technology works, then keeping up with all these new solutions coming onto the market, testing new things, trying new things, trying to suit the researcher’s question from the technological aspect, and for me personally, because we are still within the research department, being closely connected to the research and science itself. So it’s not for us just the sample coming in and the results going out, but we like and can also understand the biological relevance behind the service.

Q: What advice would you have for labs that are doing similar work and are experiencing challenges with their sample prep?

A: I think this is not advice for just the labs in my line of work, but everyone should understand the relevance of what they do. It’s not enough to follow a protocol if you don’t understand every step, because there will always be variability where you will have to adapt, and if you don’t have the knowledge, you will not be able to solve the challenges. Also, one of the most important features of our work is to have thorough quality control of our work in order to provide the best possible service.