Columbia University Launches VenomKB: The World's First Medical Venom Database

Have you heard of toxins that, rather than being lethal, can actually heal? Driven by growing interest in the therapeutic value of animal venoms, a pair of data scientists at Columbia University has created the first catalog of known animal toxins and their physiological effects on humans.

VenomKB, short for the Venom Knowledge Base, summarizes research findings from 5,117 medical literature sources describing toxins as analgesics and as therapeutic agents for conditions such as cancer, diabetes, obesity, and heart failure. Automated analysis of these publications reveals that the VenomKB database documents approximately 42,723 effects of various venoms on the human body. Although modern medicine has thus far utilized only a small fraction of the recorded toxins, researchers hope that this catalog will stimulate the discovery of novel compounds and therapies.

“Based on this list, we can take stock of the known venoms and their therapeutic effects,” said Tatonetti, an Assistant Professor of Biomedical Informatics at Columbia University Medical Center and a member of the Data Science Institute. “The question now is: How can we leverage this information in conjunction with other databases to discover new compounds and therapeutic agents?”

Tatonetti served as Romano’s thesis advisor. Together, they searched a database of 22 million medical research papers using the term “venoms/therapeutic use,” which yielded 5,117 venom-related studies. They summarized these findings using computer algorithms. After cross-referencing toxins with various drugs and correcting inconsistencies in other data, they identified 42,723 venoms with specific physiological effects. Their results were published in the journal Scientific Data, in an article associated with VenomKB.



The healing properties of venom contradict its rapid and lethal effects in the wild. Researchers have found that, across more than 173,000 species, venom has evolved over millions of years to target molecules associated with disease.

By mimicking or modifying how these toxins act on specific human cells, researchers have been able to develop drugs that suppress pain or treat diseases, often with fewer side effects than those currently available on the market.

According to a 2013 report by National Geographic: To date, researchers have developed more than ten major drugs based on the aforementioned strategies. The first to be developed was an anticoagulant named Arvin, which gained prominence after a physician discovered snake venom proteases in the late 1960s. Arvin is a purified component of Malayan pit viper venom and is used as an anticoagulant in thromboembolic diseases for the treatment of blood clots in the legs.

Byetta, a medication widely used in type 2 diabetes, is derived from exenatide, a toxin found in the saliva of Gila monsters native to the United States and Mexico.

Another drug, bombesin, which is derived from the toxin in the skin of the European fire-bellied toad, can treat gastrointestinal disorders.

Researchers conducted clinical trials on five compounds produced by the venomous cone snail, including ziconotide, a morphine-like analgesic.

Of the 5,117 publications in VenomKB, approximately 18% focus on Malayan pit vipers, Gila monsters, European fire-bellied toads, and cone snails. However, millions more venomous species remain unstudied. Zoltan Takacs, a toxicologist and Ph.D. in evolutionary studies from Columbia University, estimates that there are still 20 million toxins awaiting screening.

VenomKB was established during Romano and Tatonetti’s research on the pharmaceutical applications of venoms. Their advisor suggested that Romano look for a venom database. To their surprise, they could not find even a single venom catalog, let alone a comprehensive database. Consequently, they took the initiative to create their own venom database, VenomKB. Although VenomKB is currently modest in scale, it will become increasingly valuable as researchers contribute more data. Romano stated, “With a sufficiently large database, we can screen for compounds that are more effective and safer.”

Databases of compounds and their biological effects have been used in recent years to discover and develop new drugs, as well as to identify issues with currently marketed medications. Tatonetti and his colleagues mined a federated database of drug side effects—the FDA Adverse Event Reporting System (FAERS)—and discovered that the interaction between the antidepressant paroxetine, sold under the brand name Paxil, and Pravachol raises blood glucose levels in patients with diabetes.

With VenomKB up and running, Tatonetti and Romano plan to contribute their own data. Starting with dried venom samples from the black mamba, they will experiment with and explore new treatments for chronic pain, diabetes, and heart disease.

Source: Columbia University

Translation | Chen Kun

Editor: Huang Jia