Abstract by Ahmed Barakat

Since the first use of opium poppy 3000 BC, analgesic drug discovery has relied mainly on phenotypic screening approach, where many currently approved drugs such as opioids, non-steroidal anti-inflammatory drugs (NSAIDs), local anesthetics were discovered based on their phenotypic activity in patients without much knowledge about the underlying mechanism of action. With the molecular biology revolution in the eighties, drug discovery has shifted towards target-based drug discovery approach. Following this approach, a wide variety of molecular assays have been developed and applied in pain research e.g., knockout animal models, fluorescence calcium imaging. Despite being successful in other research areas such as cancer and autoimmune research, this approach did not contribute much to new analgesic approvals. This has been attributed to a variety of reasons including the complex, multifactorial, multiscale, multisite nature of pain and the translational gap between species. Another area of research methodology (data science) is recently adopted in pain research. This new methodology is based on acquiring and analyzing small- or large-scale data to gain novel and unbiased insights. Application of this approach i.e., data science to pharmacological research i.e., computational pharmacology is showing promises to better understand disease mechanisms and design more effective and safer drugs for different diseases. The first chapter of this thesis “General introduction” provides a discussion of computational pain pharmacology approach including theory, tools, advantages, limitations, and applications to meet drug discovery challenges. The second “Study #1” and third “Study #2” chapters describe application of computational pharmacology to address two current challenges in pain research: translational potential and validity of animal behavioral tests; and understanding mechanisms of multiple myeloma bone pain and prioritizing therapeutic targets for effective management.