Uniphage

Antibiotics are traditionally used to combat bacterial infections. However, they are becoming increasingly ineffective while the total global impact of bacterial diseases continues growing, already reaching almost 300 billion USD annually. Bacteriophages, safe viruses infecting bacteria only, are one of the most promising alternatives. They kill only target bacterial pathogens, unlike antibiotics that eliminate most bacteria around, including the beneficial ones. However, very few bacteriophage solutions are currently available, primarily because the current, century-old method to produce bacteriophage solutions is very inefficient: it is outdated, time-consuming, and non-scalable. Uniphage uses deep learning to substitute this manual phage selection process by computationally predicting the best phages to eliminate target bacteria. Uniphage’s technology will soon allow for production and testing of new antibacterial solutions against any bacterial pathogen within weeks. Using this significant improvement, Uniphage aims to become the go-to solution to fight bacterial diseases across industries. 

The current mainstream solution to tackle bacterial infections - antibiotics - is becoming increasingly ineffective and forbidden from usage outside of human health. Moreover, antibiotics eliminate most bacteria, including the beneficial ones, thus severely damaging fragile ecosystems and causing long-term health consequences. Phages, viruses naturally killing bacteria only, are one of the most promising alternatives to tackles bacterial infections. Phages are already successfully used for a small number of commercial non-human health applications, such as plant disease control and food preservation. However, only few commercial phage applications exist because the current method to produce phage-based solutions is very manual, time-consuming, and ineffective. Revolutionizing this phage selection and production process is a key to bringing more bacteriophage-based solutions to the market. This development will not only allow to significantly reduce the rate of infectious diseases, thus saving agricultural products from diseases, but also significantly reduce antibiotic & pesticide usage. 

Uniphage uses deep learning (state-of-the-art natural language processing) to eliminate the key bottleneck in bacteriophage production - the century-old bacteriophage selection method. Uniphage has developed the most efficient models to computationally predict phages against target bacterial pathogens, thus making it possible to produce new antibacterial solutions within mere week. Uniphage will sell its solutions both in the B2B and B2C manner, with individual consumers applying the solutions according to easy instructions. Uniphage's very first target is citrus greening, a currently insurable and devastating disease that causes the US $2B+ in damages annually. The solution against citrus greening will be delivered in a spray and/or injection form and applied by the individual citrus growers in FL, CA, and beyond.  

We’re targeting multiple SDGs, inc. SGD 3: Good Health and Well-Being, (SGD 13) Climate Action, (SDG 14) Life Below Water, and (SGD 9) Industry, Innovation, and Infrastructure, and 

SDG 2: Zero Hunger - through our work, we aim to save food from bacterial diseases. Currently, between 25-40% of all crops are lost to pathogens. Around 7-15% are lost to bacterial pathogens. At Uniphage, we aim to produce affordable antibacterial solutions for agricultural purposes (starting with curing the citrus greening disease) that will significantly reduce the burden of bacterial diseases in agriculture by at least several percentage points. On the micro-level, we expect that each farmer using Uniphage’s products for their respective crops will be able to reduce their bacterial disease burden by at least 70%. 

SGD 12: Responsible Consumption and Production and SDG 15: Life on Land - antibiotics and pesticides are currently widely used for food disease control. These compounds tend to be toxic and dangerous for humans, animals, and the environment. Our solutions, on the other hand, are bacteriophage-based, which means they are non-toxic, safe, and environmentally friendly. Therefore, through increasing our solution availability and widespread use of it, we contribute to increasingly responsible production of agricultural and livestock products as well as making life on land safer and more sustainable.

25-40% of crops are lost to pathogens annually. To combat this, we pour dangerous chemicals, inc. pesticides and antibiotics, into the environment and expand agricultural lands. Agricultural land expansion accounts for 78% of all deforestation worldwide. We aim to eliminate the key root of this issue by producing a platform technology that eliminates the currently existing bacterial diseases in agriculture & allows to easily and quickly combat future emerging ones. We’ll give control over their production practices to agricultural producers by distributing the ready-to-apply, safe and sustainable antibacterial solutions directly to agricultural producers to protect products from various bacterial diseases.

Uniphage has developed the most efficient models to date to computationally predict phages against target bacterial pathogens to replace the current phage selection process. We have also validated our market through the NSF I-Corps Bay Area Course & participated/are part of 5 competitive accelerator/incubator programs (all non-dilutive)

Uniphage is building a platform technology for the rapid development of bacteriophage solutions. Uniphage has already developed the most efficient to date deep learning models to computationally predict which phages can kill target bacterial pathogens. Currently, it takes years to select appropriate bacteriophages with a manual method, which is almost a century old and very inefficient. 

With our deep learning approach, we're making it possible to select appropriate bacteriophages within days to weeks, including the lab testing stage. Uniphage’s models can also help predict how to synthesize perfect bacteriophages de-novo and rapidly tweak phage genomes against bacterial resistance. 

Importantly, our computational approach also allows us to select bacteriophages against uncultivable pathogens, which cannot be achieved with the currently existing, lab-based methods. 

To build its models, Uniphage improves on the state-of-the-art natural language processing approach, which has never been applied to the bacteriophage field. Particularly, we treat genetic code the same way natural languages, such as English and Spanish, are traditionally treated. 

Lab-based experiments to test Uniphage’s model performance are conducted by a Canadian phage repository at the University of Laval on a contract basis.

Currently - none (regulatory approval (EPA) is required) 

1 year - none (regulatory approval (EPA) is required) 

5 years - 

farms whose agricultural products are saved from diseases (~10k)

those significantly impacted downstream (on the order of high hundreds of thousands - several millions)

Long-term - many more agricultural & livestock producers (and those buying those products) as well as those directly saved from bacterial diseases

Our current KPI's are R&D-focused, i.e, measuring model's performance by performing lab-based testing of whether the predicted phages infect the bacteria of interest. Our post-regulatory-approval KPI's will focus on #of solutions available against different bacterial diseases & #L of solution sold. We will use these KPI's because the more our solution is used, the more impact we will create.

Full-time (founders) - 2

Part-time - 2 

Intern - 1

Collaborators - Canadian Phage Bank 

Team Composition:

Sofia (CEO) graduated from Yale-NUS College with a degree in Life Sciences. She has been involved in various research projects, from biosensors to small RNAs. She was also selected as Biotech Leader of Tomorrow, Amgen Scholar, Epic Fellow at Conservation X Labs, and more. Sofia is very excited about working at the intersection of AI and Microbiology to bring unforeseeable innovations to our everyday lives. https://www.linkedin.com/in/sofiasp/

Chris (CTO) - Chris is a graduating Yale-NUS senior majoring in CS. He has extensive experience working in the data science and software engineering fields. Before joining Uniphage, Chris was taking graduate-level ML modules at Stanford University while working as a data scientist at Cuberg, a next-gen battery startup that just got acquired. Chris is passionate about entrepreneurship, innovation, and real-life applications of AI. https://www.linkedin.com/in/chrislis/

To increase our science expertise, we have scientific advisors (both professors at Yale-NUS College): Maurice Cheung, plant bioinformatician, and Nathan Harmston, computational biologist. 

Other team members come from Applied Math, Finance, and Cell Biology backgrounds. 

Founders' journey:

Chris and Sofia have been friends since April 2017, when they first met at Yale-NUS College. Sofia and Chris previous worked together on a different project, and even after making a difficult decision to abandon it, they remained friends and decided to work together again. This demonstrates that the Uniphage founding team will stick together even when facing difficulties. 

Unique:

We are among very few early pioneers bringing cutting-edge natural language processing techniques into the traditionally underfunded bacteriophage field

1. Uniphage's CEO is female 

2. All Uniphage's team members come from different countries and often continents & backgrounds 

3. Both co-founders have lived on different continents and have a significant multinational exposure, which helps them think out-of-the-box 

4. We aim to continue recruiting a very diverse set of people because we can clearly see how it benefits our team & a lot of unique ideas are born and developed 

5. Other unique facts about our team: 

a. Our cutting edge AI & biotech startup was founded by two final year undergraduate students, which makes us the first biotech & deep tech startup launched by undergraduates in a Singapore-based university

b. We are among very few early pioneers bringing cutting-edge natural language processing techniques into the traditionally underfunded bacteriophage field. This includes re-solving another key bottleneck in the whole genomics field - new, non-k-mer based genome representation approach (already ahead of any published literature).

c. We will continue resolving bottlenecks and breaking boundaries in the ML & biotech space to allow for many more powerful technologies to build off our developments 

By being part of the Boston-based Petri Community of Female Founders, Sofia came to realize the immense power of being in such a strong tech ecosystem - from partnering with leading academics to bringing talent on board, one could always find advice in that amazing group. Over time, Sofia came to strongly believe that there is no better place to build and grow a biotech, impact startup than in the MIT-Harvard ecosystem. 

Uniphage is very excited about joining the Solver community to more closely integrate and collaborate with MIT Solve and general MIT ecosystem, including R&D collaborations, founder & investor networking, and much more, especially as Uniphage considers moving its headquarter to Boston.

Uniphage also hopes to build strong personal relationships and partnerships with other Solvers to brainstorm ideas, collaborate on projects, and just push each other forward with the best intentions possible.  

We believe that we can most benefit from academic and other R&D partnerships with other Solvers as well as MIT labs. In the future, getting help in bringing the best talent on board as well connecting to investors would be invaluable.

Deep learning, bacteriophage, and other viral labs & startups at MIT (there are A LOT)

We aim to give agricultural growers (as well as other populations) from all over the globe the opportunity to protect their products from dangerous bacterial diseases in a sustainable and safe fashion. We expect to make our solutions available in a range of developed & developing countries over years to come as we progress with our R&D and developing an increasing number of bacteriophage-based solutions. 

Uniphage is led by a female CEO who deeply cares about female empowerment in deep-tech and beyond. She particularly believes that by reducing the financial burden on developing communities by reducing rates of crops killed by bacterial pathogens, females in developing countries will be able to particularly benefit through the financial gains and saved time, esp. through new educational opportunities. 

Through our deep tech solution, we aim to give agricultural growers (as well as other populations) from all over the globe the opportunity to protect their products from dangerous bacterial diseases in a sustainable and safe fashion. We expect to make our solutions available in a range of developed & developing countries over years to come as we progress with our R&D and developing an increasing number of bacteriophage-based solutions. 

Deep learning and cloud platform technologies constitute the core of our company. Through getting additional funding through The AI for Humanity Prize, we will be able to train better-performing models, which will translate into more accurate, better working solutions. This, in turn, will translate into more antibacterial solutions being available in developed and developing countries, in agriculture and beyond. 

Our dream is to build a GPT3-sized model for precise genome and viral engineering, and we will certainly use this prize to advance to our goal :) 

Deep learning and cloud platform technologies constitute the core of our company. Through getting additional funding through The GSR Prize, we will be able to train better-performing models, which will translate into more accurate, better working solutions. This, in turn, will translate into more antibacterial solutions being available in developed and developing countries, in agriculture and beyond. 

Our dream is to build a GPT3-sized model for precise genome and viral engineering, and we will certainly use this prize to advance to our goal :)