Latest Posts:

Sorry, no posts matched your criteria.

Follow Us:

Back To Top
DSV

Computational Drug Discovery

The next wave of innovation in pharma

Our ability to influence biology is reaching a point at which the multidimensional nature of biological data (e.g. complex off-target effects, heterogeneity of cells, temporal and spatial dynamics, resistance and feedforward / feedback networks) is slipping beyond human comprehension. In order to see a significant acceleration in the identification of effective cures for many of our most challenging diseases we need to take a far more data-driven approach. Despite several false starts big pharma is actively searching for AI solutions that improve productivity and enable new solutions, together we can accelerate this transition.

"We find it unbelievable that a 12% successes rate is considered acceptable in therapeutic design."

The next wave of innovation in pharma

Our ability to influence biology is reaching a point at which the multidimensional nature of biological data (e.g. complex off-target effects, heterogeneity of cells, temporal and spatial dynamics, resistance and feedforward / feedback networks) is slipping beyond human comprehension. In order to see a significant acceleration in the identification of effective cures for many of our most challenging diseases we need to take a far more data-driven approach. Despite several false starts big pharma is actively searching for AI solutions that improve productivity and enable new solutions, together we can accelerate this transition.

"We find it unbelievable that a 12% successes rate is considered acceptable in therapeutic design."

The next wave of innovation in pharma

Our ability to influence biology is reaching a point at which the multidimensional nature of biological data (e.g. complex off-target effects, heterogeneity of cells, temporal and spatial dynamics, resistance and feedforward / feedback networks) is slipping beyond human comprehension. In order to see a significant acceleration in the identification of effective cures for many of our most challenging diseases we need to take a far more data-driven approach. Despite several false starts big pharma is actively searching for AI solutions that improve productivity and enable new solutions, together we can accelerate this transition.

"Smart convergent approaches around bioinformatics, multi-omics, AI, etc. offer the opportunity to drive the creation of vastly more effective and personalised therapies."

Why we need to work together in this space

Building a digital company in therapeutics is hard. Unfortunately several overly ambitious early efforts delivered very little and this has driven scepticism sky high both within pharma and across the traditional investor base.
– There is limited visibility of the specific opportunities and needs for applied AI across the Pharma value chain, generating a mismatch between the problems that exist and the solutions which are created.
– Business models are challenging and, unlike traditional enterprise, come with complex and ever changing dynamics around ways to build partnerships and value product.
– Defensibility can be extremely challenging as the data is often owned by the partner.
– Ventures in this space often struggle to access capital from either traditional biotech investors or digital investors.
– Accessing data (or facilities to produce data) is often extremely challenging as is accessing talent.

Why we need to work together in this space

Building a digital company in therapeutics is hard. Unfortunately several overly ambitious early efforts delivered very little and this has driven scepticism sky high both within pharma and across the traditional investor base.
– There is limited visibility of the specific opportunities and needs for applied AI across the Pharma value chain, generating a mismatch between the problems that exist and the solutions which are created.
– Business models are challenging and, unlike traditional enterprise, come with complex and ever changing dynamics around ways to build partnerships and value product.
– Defensibility can be extremely challenging as the data is often owned by the partner.
– Ventures in this space often struggle to access capital from either traditional biotech investors or digital investors.
– Accessing data (or facilities to produce data) is often extremely challenging as is accessing talent.

"Smart convergent approaches around bioinformatics, multi-omics, AI, etc. offer the opportunity to drive the creation of vastly more effective and personalised therapies."

Why we need to work together in this space

Building a digital company in therapeutics is hard. Unfortunately several overly ambitious early efforts delivered very little and this has driven scepticism sky high both within pharma and across the traditional investor base.
– There is limited visibility of the specific opportunities and needs for applied AI across the Pharma value chain, generating a mismatch between the problems that exist and the solutions which are created.
– Business models are challenging and, unlike traditional enterprise, come with complex and ever changing dynamics around ways to build partnerships and value product.
– Defensibility can be extremely challenging as the data is often owned by the partner.
– Ventures in this space often struggle to access capital from either traditional biotech investors or digital investors.
– Accessing data (or facilities to produce data) is often extremely challenging as is accessing talent.

challenges

Better targets

Firstly, there is an increasing need for multi-omic platforms to interpret the impact of spatio-temporal dynamics of disease and associated characteristics including; evolution of genetic signature, microbial communities, cellular heterogeneity, transcription, translation, signalling, metabolism and the complex relationship between enhancers and promoters and their resulting action.
Secondly, it will be essential to better integrate existing knowledge into this multi-dimensional understanding both by leveraging internal data (including repositioning potential) and using data structures which effectively map this vast web of relationships in an easily explored manner.

Drug design and optimisation

New targets will require new approaches. Firstly, this will require advancements in; targeting and positioning based on specific pathological subsets of cells via multi-dimensional genetic, epigenetic transcriptomic, proteomic, morphological, metabolic signatures. Secondly, advanced protein design and prediction of interaction including specific focus areas such as MHC binding optimisation. Thirdly, small molecule design, diversity and improved search paradigms (CADD v2). The final piece of the picture is of course advanced delivery platforms.

Trials, treatment and disease management

With 80% of patients for the top 20% of prescription drugs being non-responders and only 5-10% of drugs reaching approval, there is likely significant scope to reduce losses during trials. This includes ‘omics models, bayesian approaches to max dose and phase 0 trials, better mapping of drug interaction and bioactivity, predicting patient trajectories, long term monitoring and non-invasive diagnostics and more representative in-vitro models.

Join us to build the next wave of companies in pharma