Connect with us

Technology

10x Science Raises $4.8M Seed to Build AI That Understands Proteins at the Molecular Level

Published

on

SAN FRANCISCO BAY AREA, Calif., April 22, 2026 /PRNewswire/ — 10x Science, which builds frontier AI for molecular-level protein characterization across the life sciences, today announced the closing of its $4.8M seed round led by Initialized Capital. The oversubscribed round includes investments from Y Combinator, Civilization Ventures, Founder Factor, and a group of strategic angel investors. Starting with drug development, the company’s platform delivers automated, explainable molecular insights in minutes where current tools and manual workflows require months. With tens of thousands of biologic drugs in active development worldwide and regulatory demands for molecular characterization intensifying, 10x Science is unlocking a new category at the intersection of AI and the life sciences.

Protein characterization is foundational to drug development. Every biologic therapeutic, from cancer immunotherapies to gene therapies, must be characterized at the molecular level to determine whether it is safe, effective, and manufacturable. Today, this work depends on specialized scientists spending weeks or months manually interpreting complex mass spectrometry data using tools that have not fundamentally changed in decades.

The pharmaceutical industry is developing more complex protein therapeutics than ever before, and the demand for characterization is growing far faster than the supply of experts who are trained for it. The 10x Science platform addresses this bottleneck with a purpose-built AI architecture that reasons across hundreds of thousands of spectra, identifies molecular forms and chemical modifications, and delivers comprehensive, explainable results.

10x Science was founded by David Stephen Roberts, Ph.D., Andrew Reiter, and Vishnu Tejus, out of Professor Carolyn Bertozzi’s Nobel laureate laboratory at Stanford University. The three founders shared a common frustration: they were studying what happens molecularly when an immune cell meets a cancer cell, one of the most critical problems in cancer research, and the tools they needed to characterize the proteins involved did not exist.

Roberts, a Damon Runyon Cancer Research Fellow with over 38 publications in the Nature and ACS families of journals, spent his career developing the foundational science of next-generation protein characterization.Reiter trained at the Broad Institute of MIT and Harvard, where he built the analytical tools pharma uses to understand how drugs bind their targets, before joining the Bertozzi lab at Stanford as a Ph.D. student.Tejus is a two-time Y Combinator founder who went to college at age 11. The three founders share a conviction that the life sciences and the technology world have been working on deeply complementary problems without ever connecting, and 10x Science is the bridge between them.

“The people building AI have historically not been life scientists, and the life scientists have not been building AI; we come from both worlds,” says co-founder and CEO David Stephen Roberts. “We realized we could build something that had never existed: an AI system with the scientific depth to reason about proteins the way the best experts do, but at a speed and scale no human team can match. For the first time, we can begin to ask the question that the entire pharmaceutical industry has never been able to answer: across thousands of characterized therapeutics, what molecular patterns distinguish the drugs that work from the ones that do not.”

The platform’s core capability is deep memory: it learns from every dataset, processing and developing an increasingly deep understanding of each customer’s molecular portfolio over time. Every result is explainable and traceable, which is essential in a regulated industry where characterization results appear in filings to the FDA. Legacy tools start from zero with every analysis. Combined with the founding team’s unique expertise at the intersection of chemistry, biology, mass spectrometry, and modern AI architecture, the company is positioned to define a new category in the life sciences.

“AI has already made meaningful contributions to biology at the prediction layer, asking what a protein might look like based on its sequence,” says co-founder and COO Andrew Reiter. “What no one has built is AI for the characterization layer, where you interpret real experimental data from real therapeutic molecules: that is the layer where drug development decisions are actually made, and it has remained painfully manual.”

The company’s vision extends well beyond faster protein characterization. As the platform processes more molecules across more organizations, 10x Science is building toward a shared layer of molecular intelligence for the life sciences: a deep, evolving understanding of protein therapeutics grounded in real experimental data at a depth and scale that has never existed.

“This is a critical moment in pharma; the industry is looking for AI that actually works, and protein characterization is needed at every stage of the drug lifecycle regardless of whether any single drug succeeds or fails. We’re talking about the infrastructure layer of drug development,” says Zoe Perret, partner at Initialized Capital. “The 10x Science founders helped build this field, and they’re now showing up with a product that solves an expensive, critically important problem. There is no more credible team to do this.”

“Biologics are the fastest-growing segment of the pharmaceutical industry and are the most complex to develop. Every antibody, every cell therapy, every engineered protein requires characterization at a level of detail that existing tools simply weren’t designed to handle. The field has outgrown its infrastructure. That’s not sustainable,” says Carolyn Bertozzi, 2022 Nobel Laureate and Stanford University Professor. “I’ve spent my career at the intersection of chemistry and biology, trying to understand how molecules behave in living systems. The biggest constraint I see across the field, whether in academic labs or industry, is the gap between the data we can generate and the insights we can extract. 10x Science closes that gap.”

Right now, the pharmaceutical industry is sitting on an enormous amount of molecular knowledge that has never been aggregated or learned from at scale. 10x Science’s AI can characterize any protein, with implications spanning cancer biology, neurodegeneration, infectious disease, agriculture, and fundamental research into how living systems work. With this funding, 10x Science is hiring Founding Engineers and expanding its work with pharmaceutical and biotech partners to open the doors for these applications.

“If we build this right, we give people across every discipline access to a new paradigm of molecular understanding that has never been possible before,” says Roberts. “10x Science can be the foundational layer of molecular intelligence for the life sciences. If we are, the world gets a deeper understanding of the molecules that govern health, disease, and life itself. That understanding belongs to everyone.”

For more information, visit: https://www.10xscience.com/

About 10x Science:

10x Science was founded in December 2025 by David Stephen Roberts, Ph.D., Andrew Reiter, and Vishnu Tejus out of Professor Carolyn Bertozzi’s Nobel laureate laboratory at Stanford University to build the first AI-native protein characterization platform for the life sciences. Roberts is a Damon Runyon Cancer Research Fellow with over 38 publications in the Nature and ACS families of journals. Reiter trained at the Broad Institute of MIT and Harvard building new tools to decipher drug interactions. Tejus is a two-time Y Combinator founder who went to college at age 11. The company builds frontier AI models with deep memory that deliver automated, explainable molecular characterization of protein therapeutics, serving pharmaceutical companies, biotechs, and research institutions. 10x Science is a Y Combinator W26 company headquartered in the San Francisco Bay Area and has received $4.8M in seed funding led by Initialized Capital. 10x Science’s frontier AI models are building a new paradigm for how scientists understand biology at the molecular level, starting with drug development.

View original content:https://www.prnewswire.com/news-releases/10x-science-raises-4-8m-seed-to-build-ai-that-understands-proteins-at-the-molecular-level-302750622.html

SOURCE 10x Science

Continue Reading

Technology

AI isn’t making students lazier – instead it’s making them more paranoid according to research by Up Learn

Published

on

By

Three quarters of students and teachers find it hard to trust AI’s accuracy, while almost two thirds of students worry they’re becoming too reliant on the technology

LONDON, Sept. 10, 2026 /PRNewswire/ — The debate around AI in schools has largely centred on plagiarism and cheating, yet new research, from online learning platform Up Learn, shows both students and teachers are more concerned about accuracy and overreliance than they are about corner-cutting.

In a survey of over 2,800 students and teachers, accuracy ranked as the top AI concern for both groups – cited by 74% of students and 73% of teachers who use the technology, providing a rare point of agreement.

Cheating, however, was cited by only one in four students and sat behind student fears of becoming too reliant on AI (65%), and losing the ability to think for themselves (47%). This gap makes sense given how students are actually using AI – students were far more likely to use AI for explaining difficult concepts (80%) and creating summaries (61%) than for essay-writing support (40%). Most are turning to AI to enable learning rather than outsource it.

Four in ten students using AI (41%) said they were concerned it isn’t aligned with their exams, while teachers raised the same issue at a higher rate (48%). Among students using AI specifically to generate exam practice questions, 45% simultaneously doubted it was aligned with the exams they were preparing for.

Beyond accuracy, more than half (56%) of teachers don’t believe AI is saving them time given how much they have to check its outputs, and a third (33%) cite privacy and safeguarding concerns.

Yet both groups keep using AI at scale. Seven in ten students (70%), and teachers (69%) are actively using it, with 42% of student AI-users turning to it daily or more, and 76% of teacher AI-users drawing on it at least weekly.

If students are turning to AI to learn rather than cheat, the question is not about discipline, but whether the tool is fit for purpose.

Guy Riese, CEO and founder of Up Learn, says: “AI is solving an education gap. Students are using it to understand things they’ve been taught but haven’t grasped and they are rightly sceptical about the answers they get. This is where we need to support students – by helping them turn that scepticism into a skill: knowing when AI has got it right, and when to look again. AI is part of the new normal, students need to be equipped with tools and techniques that enable them to use it with trust.” 

For students looking to start the new school year strong, Riese suggests three questions to ask of anything they’re learning from:

Can I check it? Before using any resource for your learning, can you see where the answer came from? Was that answer built by subject experts? Both need to be yes before you rely on it.

Is it aligned to what I’m going to be examined on? General knowledge about a subject is not the same as being prepared for a specific exam specification. General-purpose tools will only provide the information you think to ask about, so it’s better to use something built specifically for the exam you’re preparing for.

Am I doing the thinking or watching it be done? Learning should feel hard – this is known as ‘desirable difficulty’ and it’s how you know it’s working. Decades of cognitive science research have shown that the best way to learn is actively. Practising knowledge retrieval through techniques like repetition is the best way to truly absorb information and set yourself up for success when exams come around.

Notes to editors

Methodology

Research was conducted by Up Learn via an online UK survey of  2,591 students and 248teachers and in its contact base between 1 May and 17 August 2026.

 About Up Learn 

Up Learn is an adaptive attainment platform for GCSE and A Level, built by teachers and educational scientists. It combines expert teaching, adaptive learning and cognitive science, with AI supporting rather than replacing learning. It is trusted by 685+ schools and used by 1 in 3 A Level students in the UK.

www.uplearn.co.uk

View original content to download multimedia:https://www.prnewswire.co.uk/news-releases/ai-isnt-making-students-lazier—instead-its-making-them-more-paranoid-according-to-research-by-up-learn-302874127.html

Continue Reading

Technology

iQIYI Advances Professional AI Filmmaking with Launch of Global AI Creator Challenge Powered by NadouPro and Alibaba’s Wan3.0

Published

on

By

BEIJING, Sept. 10, 2026 /PRNewswire/ — On September 10, iQIYI, China’s leading online entertainment platform, launched the Global AI Creator Challenge, the first such contest on the international version of NadouPro, its AI platform built for professional film and television production. Themed “Make Ideas Unforgettable”, the challenge upgrades NadouPro’s creative toolset and partners with Alibaba’s Wan3.0 model to enable seamless, end-to-end AI-driven film production. The partnership ensures that creators have access to top-tier AI capabilities, supporting every stage of the creative process from initial ideas to finished films. By working with advanced models like Wan3.0, NadouPro is reinforcing its position as a global hub for AI-driven filmmaking, encouraging global creators to explore new possibilities in professional video production.

The NadouPro x Wan3.0 AI Challenge

The contest is designed to identify and reward high-potential creative talent using professional-grade AI tools. It has no limitations on subject matter. Creators are invited to submit AI video works of any form and explore new possibilities of AI filmmaking. Key details include:

Submission requirements: Entries must be original videos between 30 seconds and five minutes in length, with at least 60% of the content generated by the Wan3.0 model and 80% on the NadouPro platform. Global content is encouraged. To submit, participants ought to publish their work on social media platform X with the tags #NadouPro and #wan3, tagging @nadou_pro and @Alibaba_Wan.

Timeline and participation: Submissions are open from September 10 to September 23, 2026, followed by a judging period from September 24 to September 30.

Featured awards: The challenge recognizes 156 winners across three distinct categories, including 3 winners for Masterpiece Award, 3 for People’s Choice Award, and 150 for Active Participation Award.

Prize pool: A total prize pool of over USD 10,000 is available. Following the contest, NadouPro and Alibaba’s Wan3.0 will promote outstanding entries to increase their visibility among a larger audience.

Empowering Global Workflows with Leading Technology

NadouPro unites leading global models with proprietary capabilities to support creators through the entire creative journey – from script and assets to storyboarding and the finished film. The international version now supports multiple languages, including English, Korean, Thai and Arabic, serving creators in more than 20 countries and regions worldwide.

Concurrently with the contest, NadouPro is rolling out permanent creative updates to enhance the entire platform:

3D Directing Studio: Stage 3D models and cameras to refine movement and sequencing via a timeline. This creates a spatial plan to lock in complex choreography and composition before final generation.

720° Panoramas, Multi-Angle and Multi-Grid Generation: New image nodes for panoramas and multi-angle views enable one-click reference libraries, helping improve visual and character consistency across production.

Canvas Agent & Skills Library (Coming Soon): An upcoming agent will generate creative plans from a single prompt and automate node-linking once the plan is confirmed by the user, while drawing on a specialized library of skills for professional-level tasks.

The Wan3.0 model powers the contest with advanced realism and storytelling capabilities. Its Omni-Reference feature supports multi-modal inputs including images, video, audio clips, and documents. Separately, it can generate native single-shot videos of 30 seconds. The model supports pixel-level consistency for referenced assets while providing an immersive audio-visual experience with synchronized sound. Creators also benefit from precision video editing through both instruction- and reference-based tools for greater production efficiency.

From a single idea to a work that endures, NadouPro invites creators around the world to push the boundaries of what AI video can be. As a platform committed to advancing the future of AI filmmaking, iQIYI will continue to put professional-grade capabilities and possibilities in the hands of creators everywhere. Follow nadou.ai for the latest updates and details on how to join the challenge.

Contact:
iQIYI Press, press@qiyi.com 

 

View original content to download multimedia:https://www.prnewswire.com/apac/news-releases/iqiyi-advances-professional-ai-filmmaking-with-launch-of-global-ai-creator-challenge-powered-by-nadoupro-and-alibabas-wan3-0–302874875.html

SOURCE iQIYI Inc.

Continue Reading

Technology

SUNLU and INSLOGIC Launch FilaDC i10 Filament Dehumidifying Cabinet for Long-Term Filament Storage

Published

on

By

The new FilaDC i10 combines active dehumidification, 10-spool storage, low energy consumption and quiet operation to help maintain a stable, low-humidity environment for filament storage.

FRANKFURT, Germany, Sept. 10, 2026 /PRNewswire/ — SUNLU and INSLOGIC are introducing the FilaDC i10 Filament Dehumidifying Cabinet, a new long-term storage solution designed to help 3D printing users protect filament from moisture between prints.

Physical Adsorption + Smart Cycling

The i10 uses a molecular sieve to adsorb moisture from the air inside the cabinet. Its smart cycling system releases collected moisture and restores adsorption capacity automatically, providing continuous humidity control between 10% and 45% RH.

In testing, PETG stored in the i10 decreased from 0.34% to 0.21% moisture content after 96 hours, while PETG exposed to open air increased from 0.35% to 0.39% over the same period.

This creates a stable, low-humidity environment for long-term filament storage without relying on continuous high-temperature drying.

10-Spool Storage, Stackable Design

The i10 stores up to 10 × 1 kg filament spools and supports spool sizes from 0.25 kg to 5 kg.

Multiple i10 units can be stacked vertically to save floor space, with stacking up to three units recommended.

A sealing gasket and three strong magnets help keep the cabinet tightly closed and maintain a stable storage environment.

Built for Long-Term Storage

The i10 consumes approximately 0.23 kWh/24h* and operates below 30 dB, making it suitable for continuous use in homes, studios, workshops, and print rooms.

Its integrated design requires no complicated installation. Users can simply set it up, load the filament, and let it maintain the storage environment.

*Tested at 25°C and 60% RH in an empty chamber. Actual results may vary.

Multiple Safety Protections

For long-term operation, the i10 combines hardware temperature protection with intelligent software monitoring. If excessive temperature is detected, the system provides a real-time warning.

Beyond filament, the i10 can also provide low-humidity storage for other moisture-sensitive items, such as camera lenses, electronic components, tools, and collectibles.

Availability

The SUNLU × INSLOGIC FilaDC i10 will be available for pre-order beginning:

September 15, 2026, at 7:00 AM UTC
Pre-order: [https://bit.ly/4gGV5lb]

About SUNLU

Founded in Zhuhai in 2013, SUNLU is a global 3D printing brand specializing in filaments, resins, and accessories. With 270+ production lines, 25 million+ products sold, and 530+ intellectual property rights, SUNLU continues to advance 3D printing through product innovation, quality, and accessibility.

View original content to download multimedia:https://www.prnewswire.com/news-releases/sunlu-and-inslogic-launch-filadc-i10-filament-dehumidifying-cabinet-for-long-term-filament-storage-302873286.html

SOURCE SUNLU

Continue Reading

Trending