Removing the trial-and-error process from depression

AI Personalized Medicine

Like Comment

Download the full PDF here.

Depression is the leading cause of disability world­wide, with many patients facing an often long and painful trial-and-error process to find the right antidepressant. This can lead to patient dissatisfaction, poor treatment adherence and increased side effects. The dawn of artificial intelligence (AI) brings huge potential to change this status quo. By applying a deep understanding of the scientific and neurological data, algorithms can be developed to analyze the accumulating real-world datasets, such as from genomic banks and electronic health records. Through identifying relevant markers, highly com­plex prediction models can be built to personalize depression treatment.

Founded by brothers Dekel and Oren Taliaz in 2015, this is exactly the neuroscientific, evidence-based approach used by the Tel Aviv-based startup, Taliaz. Going beyond existing pharmacogenomic solutions (genetic tests), its AI prediction algorithm, Predictix, enables deep data analysis of hundreds of thousands of clinical, demographic and genetic parameters involved in driving behaviors, to opti­mize depression treatment. The team is currently investigating the use of this technology platform in other psychiatric disorders, such as attention-deficit hyperactivity disorder.

“Our goal is to use a scientific, critical and analytic approach to change how we think about treatment,” said Dekel Taliaz, CEO.

Personalizing antidepressant prescriptions with Predictix

Today, physicians are bombarded with data. The Taliaz platform, Predictix, transforms these complex real-world patient data into physician-centric decision support tools. Through collecting a patient’s DNA sample, alongside their demographic and clinical history via an online questionnaire, Predictix analyzes and yields a personalized prediction report for each patient. The report empowers physicians’ clinical decisions by providing predictive information about antidepressant drugs and about which treatment is best suited to each patient’s genetic makeup and health record. On the basis of data from the largest ever prospective study on depression therapies—the USA STAR*D Study—the Predictix algorithm can pre­dict efficacy and adverse effects of current antide­pressants with up to nearly 80% accuracy.

“Predictix is about learning from the choices that are made to reduce the trial and error,” said Dekel Taliaz. “As physicians use Predictix, our algorithms learn from these new datasets to further optimize our treatment predictions. We are then packaging this science for physicians and patients in an easy-to-use, noninvasive and straightforward way. It is hardcore science, but very practical.”

In March 2019, Taliaz launched the online, medically managed Predictix service in the UK with the leading online psychiatry service, Psychiatry UK. Predictix is also being implemented in France and in the Israeli health system, following an $8 million agreement with the Israel Ministry of Health.

Contact Details
Dekel Taliaz, CEO and Cofounder
Taliaz Ltd.
Tel Aviv-Yafo, Israel
Tel: +972 77 335 2506

Go to the profile of Taliaz


Taliaz is a health analytics startup that uses a unique science-based, data-driven approach to advance treatment of psychiatric disorders. The company has innovated PREDICTIX, a proprietary, artificial-intelligence-based algorithm that analyzes hundreds of genetic, clinical and demographic combinations to help healthcare providers tailor medication to each patient’s unique genetic make-up and health record. Addressing the US$17 billion depression market, its flagship application, PREDICTIX Antidepressant, is a clinical software tool to empower doctors to better treat depression. Already available in the UK and Israel, PREDICTIX may help reduce the often and long and painful trial and error process faced by patients to find the most effective antidepressant. The technology was developed using data from the largest ever prospective study on depression therapies involving thousands of patients, the STAR*D Study. By completing a simple, non-invasive genetic test and proprietary online clinical questionnaire, doctors receive a personalized report with predictive information about antidepressant drugs and which is best suited to patients’ genetic makeup and health record. This helps doctors tailor treatment plans to increase efficacy and avoid side effects, speed up patient recovery and reduce associated health system costs.

No comments yet.