Insitro, a startup that uses machine learning to accelerate drug discovery and development, has announced that it has raised $400m in series C funding to push its mission forward.
The company, which focuses on utilising machine learning to interpret molecular biology, will use the funding to enhance its cutting-edge technology and expand its capabilities.
This funding round marks a significant milestone for Insitro and its mission to revolutionise drug discovery.
Table of Contents
ToggleInsitro raises $400m for machine learning drug discovery push
Insitro is a drug discovery and development platform that uses automated laboratory systems and Machine Learning (ML) to identify promising drugs. Founded by a team of scientists and professionals from the healthcare sector, Insitro’s mission is to change how drugs are discovered, empowered by automation and AI technologies. The company has raised $400 million in its latest funding round, which it will use to fuel expansion efforts and enhance its ML-driven platform.
The Insitro platform leverages ML algorithms to analyse information from different sources including biological data, experiment results, clinical data, and genomic information. The platform analyses this data across more than 20 drug discovery metrics to optimise the selection process of candidate treatments. This process enables researchers to identify promising candidates faster and more accurately than conventional approaches. Additionally, by incorporating constant feedback loops into the development process, Insitro’s technology helps reduce experiment costs while maintaining a high accuracy rate in candidate selection.
Insitro’s platform has already been used in dozens of pharmacology studies focusing on therapeutic areas such as oncology and infectious diseases with key partners including Gilead Sciences Inc., Merck KGaA/Pfizer Alliance, Sanofi SAS/Genzyme Corp., AstraZeneca PLC/MedImmune LLC, Regeneron Pharmaceuticals Inc., Unum Therapeutics Inc., Novartis AG/Pfizer Alliance pharmaceutical companies as well as Harvard University Medical School. Moreover, Insitro also offers its technology for sale to research institutes and academic laboratories worldwide.
What is machine learning?
Machine learning is an artificial intelligence (AI) technology that enables computers and software to ‘learn’ from data. Across many different types of applications, machine learning is used to help support progress in finding new insights, understanding user behaviour, good decision making and gaining predictive accuracy.
In drug discovery, machine learning has become a powerful enabler for better understanding complex biochemical pathways and drug-target interactions. In addition, by utilising large quantities of already available information about existing drugs and targets, experts can use machine learning algorithms to identify potential treatment candidates for areas where the current treatments have been inadequate or absent.
Machine learning requires high-quality training data and algorithms that detect patterns — most commonly supervised learning models. In supervised learning, datasets are labelled to predict outcomes from new input data. This can be applied to drug discovery through Bayesian optimization or deep neural network models that use known structures of target molecules for more accurate predictions on potential drug targets. Through iterative cycles of model optimization and self-learning, the accuracy of predictions can be further improved and bring powerful new insights into systems biology research.
Insitro Raises $400 Million for Machine Learning Push
On April 27th 2021, Insitro announced it raised $400 million in a funding round that valued the company at $4.4 billion. This funding will help Insitro further develop the machine learning-enabled drug discovery platform it has been working on since its launch in 2019.
Through this platform, Insitro is looking to majorly impact drug development by leveraging its tools and data to accelerate the process. Let’s look deeper at the details of this new funding round and what it means for Insitro.
Overview of Insitro’s funding round
Insitro Inc., a San Francisco-based biotechnology company, recently closed a funding round worth $400 million. This key financial backing has enabled the firm to revolutionise its machine learning drug discovery platform and solidify its position as one of the leading firms in drug discovery innovation.
The strategic opportunities created through Insitro’s $400 million Series B raise will allow it to continue creating tools and platforms to interpret complex biological data and further accelerate drug discovery and development. The funding round was led by GV and argues for the major faith capital investors have in the capabilities of Insightro’s sophisticated machine learning platform. The funds will also enable Insitro to build their data-driven infrastructure, expanding their vision for biomedical engineering within personalised healthcare.
With the raised funds, Insitro plans to strengthen its machine learning-based ability to predict molecular properties, study complex biological systems, design new therapeutics while improving existing treatments, and facilitate process optimization techniques to accelerate drug research and development pipelines.
What will the funds be used for?
The $400 million in funding will be used by Insitro to further develop its machine learning-driven platform for drug discovery. Insitro has developed a technology which combines innovative computing techniques, human biology and machine learning tools to accelerate the development of new treatments for diseases with unmet medical needs. With this funding, the company intends to build on its existing research and development efforts to further expand its technology and capabilities.
One potential use of these funds is furthering Insitro’s focus on personalised medicine. This treatment involves developing targeted therapies to treat an individual’s medical condition. Using Machine Learning, the company can develop customised treatments based on data from patient samples collected during clinical trials. It can also use Machine Learning algorithms to predict which combination of drug components together could create a more effective treatment for an individual’s specific disease or disorder, ultimately providing more precise and powerful drugs than ever before.
The funds will also be used to expand Insitro’s operations globally. The company plans to extend its platform into other countries including China and South Korea and build partnerships with pharmaceutical companies in other territories including Europe, Japan, Canada and Australia to reach even more patients with the medications it develops most effectively. Additionally, the funds can be used to recruit top talent to increase its research capabilities and substantially invest in IT infrastructure including cloud computing resources needed for processing large amounts of environmental data crucial for machine learning efforts.
Insitro’s Platform
Insitro, a San Francisco-based biotechnology firm, recently raised a massive $400 million from investors which will help them to fund their machine learning drug discovery platform.
The platform is built to utilise the power of both data and machine learning algorithms for more efficient, accurate and cost-effective biomedicine research.
Let’s find out more about Insitro’s platform and how it works.
Overview of Insitro’s platform
Insitro is a biotechnology company that focuses on using machine learning for drug discovery through their innovative platform. Insitro’s platform uses automated, end-to-end simulations to engineer and analyse large datasets to develop computer models that accurately simulate biology and drug response in the body. This approach has significantly reduced the time and cost associated with drug discovery.
Insitro’s platform accesses vast amounts of data such as chemical library compounds, protein sequencing, gene expression, patient health records, phenotypic readings, etc., stores these assays cleanly into a proprietary database and runs probabilistic searches or simulations to identify new relationships within biological systems. The platform then combines existing biological knowledge with deep learning algorithms for predictive models used to design candidate molecules for further testing . This data-driven approach ultimately enables Insitro AI experts to develop comprehensive models of disease states or treatments faster than traditional manual methods.
In addition, Insitro has developed its proprietary manufacturing processes allowing it to rapidly produce batches of small molecules from its platform for further evaluation or use in clinical trials. With this important milestone now accessible through its machine learning-based drug discovery programme , the potential for Insitro’s new platform is immense. In April 2021 the company raised $400M in Series C funding signalling strong investor confidence in their technology platforms creating rapid innovation in the drug discovery field moving forward.
How does it work?
Insitro’s machine learning drug discovery platform is an artificial intelligence (AI) technology designed to accelerate the drug discovery process – the search for new treatments and cures. The platform utilises AI-driven technologies such as natural language processing and machine learning to analyse vast biological data and suggest novel therapeutics targets. It also incorporates multiple data sources, including Electronic Health Records (EHR), genomic information, medical literature, clinical trial results, and other proprietary datasets to draw insights that assist drug developers in the early stages of therapeutic development.
At its core, Insitro’s platform leverages massive amounts of data pulled from numerous sources to identify potential mechanisms for treating disease. Once key targets are identified, Insitro builds computationally predicted models of what may work best from a molecular biology standpoint. These models are then tested against thousands of candidate molecules in silico (virtual) experiments to help identify candidates potentially impacting a particular target at greater efficacy than existing therapies. Consequently, this approach accelerates lead molecule selection for drug development programs – saving time, money and resources for drug developers compared to traditional approaches without sacrificing quality or accuracy.
Benefits of Machine Learning for Drug Discovery
With machine learning playing an increasingly important role in drug discovery, Insitro, a startup backed by some of the biggest names in venture capital, recently announced a $400 million financing round to further its development of the first machine learning-driven drug discovery platform to target a wide range of diseases.
Let’s dive into the benefits of leveraging machine learning for drug discovery and explore how Insitro’s platform can help define a new era of drug development.
Accelerated drug development process
The combination of machine learning and drug discovery is revolutionising drug development. With machine learning systems, drug designers have access to a powerful and accelerated approach for generating insights about the potential behaviour of new compounds. In addition, compared to traditional methods, predictive models allow for earlier detection of failure points in course development – saving considerable time and resources in bringing a new drug to market.
Machine learning applications can automate the detection of, analysis of, and response to patterns in diverse data sets from literature, experimental results and clinical trials with greater accuracy and depth than previously achievable. Insitro’s proprietary platform then hastens the testing pace by using various algorithms – such as self-supervised DL networks – that autonomously determine what experiments should be done next to accelerate drug development.
By applying this powerful combination, Insitro can more quickly connect promising compounds with the right receptors and therapies while reducing preclinical timelines significantly — helping bring medicines more quickly to those who need them.
Improved accuracy of drug predictions
Machine learning is transforming drug discovery processes by accelerating lead candidate identification. By combining and analysing large datasets such as disease genomics data and patient health records with machine learning algorithms, researchers can discover correlations and patterns that are otherwise complex to discern. This enables enhanced accuracy of predictions, leading to the rapid identification of novel drug targets and improved strategies for clinical trial design.
For example, Insitro’s machine learning-driven approach uses innovative techniques such as reinforcement learning to leverage genomic data from hundreds of thousands of patients to accurately identify molecules with high potential as therapeutic agents. Further, advanced visualisation techniques are employed to observe vast amounts of generated data from target modality to identify issues or analyse trends that may affect accuracy of predictions. This approach has enabled more accurate drug prediction than traditional methods, ensuring optimal efficacy with fewer clinical trial failures associated with incorrect prediction of drug targets.
Challenges of Machine Learning for Drug Discovery
Recent advances in machine learning (ML) have enabled pharmaceutical companies to use automated processes for drug discovery. Insitro, a Silicon Valley startup, recently raised $400 million to continue pushing the boundaries of ML-driven drug discovery.
Nonetheless, numerous challenges remain to be addressed to take full advantage of ML-driven drug discovery. In this article, we will closely examine some of the obstacles facing drug discovery using machine learning.
Cost of data collection and storage
The data collection and storage cost has been a significant challenge for developing Insitro’s machine learning drug discovery platform. Before machine learning can identify new drugs, much structured data must be collected and stored as molecular structures, genomic profiling, cell assays, and other biological datasets.
In collecting this data from multiple sources such as biobanks and sequence databases and across different types of studies, it is essential to have dedicated resources for data cleaning and standardisation for successful integration.
Data storage costs can be onerous for big pharma companies who need to collect huge volumes of highly complex datasets. This rises exponentially when considering pooled clinical trials, electronic medical records (EMRs), laboratory testing records or patient surveys required to assess the safety, tolerability or efficacy profiles of newly identified drug molecules. Moreover, pre-labeled datasets are typically expensive due to the requirement for extensive labelling effort.
With rising costs associated with healthcare in many countries worldwide and increasing competition in the field of drug discovery by academic institutions as well as public research institutes, cost-efficient strategies must involve solutions that improve access to data sources as well as reduce processing costs associated with collecting and preparing them into usable formats such as feature vectors in the Insitro ML platform.
Complexity of algorithms
The complexity of algorithms used in machine learning for drug discovery can present a challenge regarding how well the algorithm can detect patterns in the data and accurately classify new compounds. Algorithms that are too simple may lead to inaccurate predictions. At the same time, algorithms that are too complex may not be able to effectively predict outcomes or consider all possible variables. Additionally, an algorithm’s complexity could lead to scalability challenges; as more data is added, the algorithm must be adapted accordingly.
Algorithm selection can depend significantly on the application for which it will be used; for example, a deep learning approach such as a Convolutional Neural Network could be employed for image analysis, while a more complex algorithm such as Support Vector Machines might yield better results when dealing with large datasets. Therefore, it is important to carefully assess what type of problem needs to be solved and select an appropriate algorithm accordingly. Furthermore, computational power and time constraints must be considered when designing ML-driven drug discovery systems.
Security and privacy concerns
As machine learning is widely used in drug discovery, underlying security and privacy concerns become increasingly prominent. Security concerns revolve around the safety of patient data and software or hardware that may be used. This data could be exposed to cyber-attacks or fall into the wrong hands if not stored properly. Privacy concerns include protecting patient data against being mined for insights or misused for personal gain. Drug researchers using machine learning datasets should also take extra steps to ensure privacy regulations are met, especially in highly regulated markets such as Europe and North America.
Insitro’s machine learning drug discovery platform claims to help identify novel treatments by combining biomedical data with certain inferences from millions of secure computing resources. To ensure the safety of patient data and adhere to security protocols, each platform offers different levels of control access – from general use to controlled use – with varying degrees of identity validation, authentication requirement, encryption protocols for classifying data sets and more stringent requirements for cloud-enabled applications that employs advanced AI algorithms such as Natural Language Processing (NLP) and deep learning.
Additionally, Insitro’s drug discovery platforms feature built-in firewall protection and proactive monitoring software that will detect threats and patch vulnerabilities automatically on multiple levels – making it secure enough to fight cyber-attacks while keeping patient privacy intact.