Amit Mishra, Associate Scientist II, Automation, is enhancing ALS TDI’s efficiency and accelerating drug discovery through 3D printing and automation.

At the end of Amit Mishra’s freshman year at the Rochester Institute of Technology, he found a 3D printer in a pile of recycling as he was moving out of his dorm. It was a moment that would later shape the direction of his career. 

“I picked it up just because it looked cool,” he remembers. “But it was broken. So, I slowly started repairing it and learning how it worked. I think that's what kicked off my serious interest in programming and technology.”

For years, tinkering with technology was just a hobby for Mishra. He studied biomedical science in undergrad, initially setting out on a pre-med track before pivoting to research. When he came to the ALS Therapy Development Institute (ALS TDI), he was hired to help our Cell Biology Team conduct tissue culture-based experiments. However, the opportunity soon arose for him to apply the skills he had developed through his love of tech.

Falling into Automation

Around the time he started, ALS TDI’s Director of Cell Biology, Dr. Kyle Denton, was exploring the use of a new software tool to detect mutations in the C9orf72 gene—the most common cause of familial ALS—in tissue samples. Previously, these samples were sent to an outside lab for analysis. Working with Dr. Denton, Mishra was able to utilize his programming experience to create an internal automated process to analyze digital data generated from whole-genome sequencing of people with ALS. 

His script helped reduce the time required for this process from weeks to a matter of minutes. Mishra says it was the first step toward defining a new role for himself in the lab.

“Slowly, I started working on more software and robotics,” he says. “Now, my role is to help scientists scale up the processes and protocols they’re doing in the lab manually through automation.”

Automation at ALS TDI

Automated systems play a crucial role in drug discovery laboratories like ALS TDI. The science of preclinical drug discovery involves extremely large numbers of small, repetitive tasks. When testing potential drugs in cellular models of ALS, for example, researchers might run assays—tests carried out in the laboratory to determine the presence, quantity, or activity of a specific substance—in hundreds or even thousands of different small tissue cultures in a single day. These assays generate even larger numbers of datapoints, which then must be organized and analyzed to generate results.

For scientists to individually complete all these tasks by hand would take an inordinate amount of time. Automated systems, however, can be programmed to do the same work that might have taken weeks, months, or longer, in hours or days. 

Mishra says that automated science at ALS TDI can be divided into two broad categories. There are software programs, like the C9orf72 screening pipeline he built with Dr. Denton, for analyzing data, as well as robotic systems that carry out physical experiments. 

Robotics at ALS TDI

ALS TDI’s robotics systems can carry out “high-throughput screening,” or automated assays that can analyze hundreds or thousands of samples in a single day. For these assays, samples of cells are grown in plates containing dozens or hundreds of small wells. The plates are then passed along through several different individual robotic devices, programmed to work in concert to carry out a variety of assays. 

Robotic devices that might be utilized for this type of assay include:

  • An automated incubator that stores cell plates at the proper temperature in a sterile environment and moves them into position for experiments.
  • A storage device that prepares chemicals for each assay and keeps them at the ready.
  • A liquid handler that aspirates, or sucks up, chemicals for the assay into small pipettes and then delivers those chemicals into the wells of a plate.
  • Devices that remove and replace lids on plates before and after chemicals are added.
  • An automated microscope that observes the results of each assay in each individual well.
  • Several computers that control the function of these various devices, record the results of experiments, and output data for analysis. 

A typical assay the cell biology team might use this system for is measuring neurofilament light chain (NfL) levels in samples of motor neurons derived from induced Pluripotent Stem Cells (iPSC). NfL is a structural component of motor neuron cells that is shed as the cell degrades over time, and is one of the most well-known biomarkers for ALS. 

“To carry out that assay, we grow motor neuron cells in a 96-well plate,” says Mishra, “And then, to put it succinctly, we add a bunch of chemicals to each well. Those chemicals bind to the proteins or molecules within those wells and, depending on what kind of conditions we've exposed those cells to, they glow at different brightnesses based on the neurofilament levels, allowing us to measure them with a microscope.”

Cor2D2: ALS TDI’s Robot Scientist

Most of the robotic devices at ALS TDI are organized in a large system known affectionately as Cor2D2, a product of years of effort and experimentation by ALS TDI scientists in collaboration with outside robotics companies. Cor2D2 had already been in service for several years when Mishra came to ALS TDI, but he says that he has grown to know it well as an automation expert. He has helped incorporate new devices and found new uses for components that already existed. His 3D printing skills have also proven useful in his robotics work. Mishra says he is currently working on prototype components that will improve the function of Cor2D2, and he has already printed custom parts that enable specialized assays utilizing another robotic device in the lab. 

“Picking things up was a slow process, because ALS TDI’s system is unique,” he says. “Because we’re a nonprofit, we can’t always afford the latest and greatest technology. We’ve had to be creative and come up with a lot of workarounds. It was really interesting to learn how Dr. Denton used these tools in the lab. That’s influenced my mindset going forward—not just using these machines as they were designed, but figuring out how to make them suit our purposes.”

Software Automation at ALS TDI

Mishra has also helped establish several entirely new automated software systems since coming to ALS TDI. Much like robotics can save time by speeding up repetitive physical tasks, these software systems help researchers automate some of the tedious work involved in analyzing data. 

For example, Mishra has worked to set up image analysis pipelines to interpret the results of the high-throughput assays conducted with the Cor2D2 system. The data generated from these assays come in the form of images, captured by robotic microscopes. These images require careful analysis to generate usable results. To run these analyses, the raw data from the captured images must be organized and formatted in a particular way. By writing scripts that can output data in usable forms, Mishra saves his colleagues hours of work.

“If you think about an experiment as going from an idea to results, my tools come in the part right before results,” says Mishra. “They take raw data and convert it into processed data that a scientist could then put into a graphing software to generate those beautiful charts we see on posters and in research papers.”

Mishra says he has created dozens of these time-saving programs tailored to the needs of his colleagues.

“The process starts with a scientist coming to me and describing what they need automated,” he says. “I ask them questions like ‘What do you want to be able to input?’ ‘What do you want as the output?’ ‘Can you show me some examples?’ And then I do my best to write a script that gets them from the inputs to the outputs with as minimal effort as possible.”

The process of creating each automated script requires a sizable investment of time and effort from Mishra, but pays dividends in time saved for his teammates.

“If the task takes a scientist five hours, for example, it'll probably take me about a week to automate it,” he says. “But it’s worth it in the end, because we might use that pipeline for several different assays down the line, and every time you use it once it’s been created, you save five more hours.”

To learn more about ALS TDI and our research to end ALS, click here.

What to Do Next: