The world of medical and biological research generates tidal waves of data and finding a way to analyze the relentless stream of bits is an unavoidable challenge for researchers. The lab scientists know their way around reagents and pipettes, but often they don’t know much programming.
LatchBio wants to save them. Today, the company announced a series A round raising $28 million to continue to build out and improve its web-based data processing platform. Its goal is to give researchers a simple, integrated tool that can analyze the data without hiring a computer scientist.
In the last few years, several companies have begun to offer low-code or no-code options that simplify many of the common steps for turning raw data into tables and charts that offer insights. They include pre-coded routines that can often be joined in a visual, drag-and-drop interface.
The trend is common across many industries. Some companies like Snowflake or Knoema offer general platforms for working with all forms of data. Others specialize. Many companies that are creating what they call “Data Management Platforms,” for instance, want to target digital marketing efforts by creating aggregate models of consumer behavior. Its tools are highly customizable for supporting the work of advertisers and marketing teams.
LatchBio’s tools are focused on the type of data generated in biology labs. It deploys pre-coded routines for unpacking and analyzing common data formats like DNA sequencing or protein folding files. The system, for example, has well-tested integration routines that work with other common tools including RNA-Seq, CRISPResso2 and AlphaFold.
“We remain obsessively focused on building a platform that hosts the best biocomputing workflows and other tools that scientists love to use and that enable them to accelerate scientific progress,” said Alfredo Andere, cofounder and CEO of LatchBio.
Before LatchBio, Andere worked at Good and Facebook, studying machine learning and data management.
Part of the LatchBio’s game plan is to bundle storage and curation with the analysis. The platform runs in the cloud, saving researchers and their institutions the need to maintain servers. As an incentive, LatchBio offers unlimited storage that’s already HIPAA and Soc-2 compliant. The service, which is billed by storage volume rests on Amazon Web Services (AWS), which LatchBio resells.
“We actually have a thing where we’re completely free for academics forever,” said Andere. “We’re committed to making this possible by subsidizing it with industry contracts.”
‘Seemingly infinite’ cloud potential
The startup is leveraging the seemingly infinite ability of the cloud to store data, a process that is an endlessly growing challenge for bench scientists. In the past, an experiment might generate some jottings in a lab notebook. Now each run of a sequencing machine can produce hundreds of thousands of lines of data spelling out the base pairs in a molecule.
“The datasets generated by today’s scientific experiments are growing at an exponential rate, but many biologists lack the computational tools they need to rapidly analyze and iterate using the results of their experiments,” said Coatue general partner David Cahn, one of the investors leading this current round. “Coatue looks forward to supporting LatchBio’s ambition to create the necessary infrastructure to hasten the pace of discovery and translational research – and importantly, help get more life-saving therapies into the clinic.”
While LatchBio emphasizes its no-code options to bench scientists, it still offers a full API and SDK for developers who may need to write custom code. Indeed, the company imagines that developers may just upload its software, also known as “workflows.” Currently, half of the workflows available are by Latch.bio staffers and half come from the community.
The workflows are generally written in Python, the language that remains the dominant programming language in biological sciences. LatchBio is adding extra structure to its coding environment to encourage data quality. Some parts of the process will enforce type safety to ensure that the wrong type of data doesn’t end up in the wrong field.
The future of biotech
Much of the current collection focuses on the fat tail of genomics research. It offers customized workflows for common tasks from gene therapy, cell therapy and research into SARS-COV-2.
“We want to create the best software engineering team in all of biotech,” explained Andere. “ I believe there’s a huge opportunity to be all the people that are tired of optimizing financial models or optimizing advertisements. They want to look for new meaningful work within that where they can still develop into the best software engineers possible.”
Andere says that the proceeds from the funding round will be used to greatly expand the staff and create better data curation and more workflows that are optimized for biologists.
“You have a chicken and egg problem because why are biologists going to use our platform if there’s no workflows to analyze their data?” asked Andere. “And then, why are computational people gonna upload workflows to analyze their data if there’s no biologist to use them? The answer is, anytime you have a chicken egg problem, just buy a chicken, right?”
“Since leading LatchBio’s seed round, we have been impressed by the dedication, focus and speed with which the team is building solutions that will make a meaningful impact on the way science is done in the lab every day,” said Lux Capital Partner Brandon Reeves. “We are pleased to once again back this team of founders and their vision to address some of the computational and logistical challenges facing those working at the cutting edge of science.”
The $28 million series A round was co-led by Coatue and existing investor Lux Capital. Hummingbird Ventures, Caffeinated Capital and existing investors Haystack and Fifty Years joined in.
VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn more about membership.
Source: Read Full Article