Machine learning is a subfield of artificial intelligence that focuses on the development of algorithms and statistical models that enable computer systems to learn from data without being explicitly programmed. In other words, it involves the use of computational methods to automatically discover patterns in data and make predictions or decisions based on those patterns. Machine learning has become increasingly important in recent years due to the explosion of digital data and the need for automated decision-making in various industries, such as finance, healthcare, and marketing.
There are three main types of machine learning:
- Supervised learning,
- unsupervised learning
- Reinforcement learning.
Reinforcement learning differs from supervised learning in that it does not need the presence of labeled input/output pairings or the explicit correction of suboptimal behaviors. The emphasis instead is on striking a balance between exploration (of the undiscovered region) and utilization (of existing information). Because many reinforcement learning algorithms for this context use dynamic programming approaches, the environment is generally represented in the form of a Markov decision process (MDP). The primary distinction between classical dynamic programming methods and reinforcement learning algorithms is that the latter does not require knowledge of an exact mathematical model of the MDP and is designed to target big MDPs where accurate approaches become infeasible.
The available machine-learning models for predicting chemical bioactivity are drawn from the databases of the medicinal classes displayed in the mini cards below.
EP2,4 receptor antagonists310EGP Add to cart
Phosphodiesterase 5 (PDE5) inhibitors465EGP Add to cart
BACE2 inhibitors310EGP Add to cart
BTK inhibitors310EGP Add to cart
CDK9 inhibitors620EGP Add to cart
Bromodomain containing protein 4 (BRD4) inhibitors620EGP Add to cart
Chemokine receptor (CC) antagonists465EGP Add to cart
Toll-like receptor (TLR7,8) agonists620EGP Add to cart
Histone Deacetylase 6 (HDAC6) inhibitors310EGP Add to cart
- The models have an accuracy of 85-95%.
- The pricing given on each model is for predicting ONLY one compound’s bioactivity.
- Based on the nanomolar potency, the models are designed to predict chemical bioactivity. If the chemical properties of the test molecule do not match the model’s feature scheme, or if the machine predicts micromolar potency, the compound will be inactive.
- The compound’s smiles code is used for data entry to forecast activity.
- After the compound(s) prediction fees are submitted to drug discovery pro, the requestor should proceed to the service request button to complete the appropriate information and upload a pdf file with the smiles code(s) of the test compound(s). The results are then forwarded immediately to the requestor via the email address provided.