Machine Learning 101: How to train your first QSAR model
My hope is that these posts will be of interest to people who want to understand more of the nuts…
How do I know if Optibrium’s predictive models work?
Data curation for model building A model can only be as good as the data it has been trained on.…
How can I make the most of my predictive models for drug discovery?
What’s the purpose of a predictive model? What’s the value of predictive models for drug discovery? Most of the undergraduate…
What’s the difference between QSAR and imputation predictive models – which method should I use and when?
We’re often asked, “What’s the difference between QSAR and imputation models?”, so I’m going to explain how the methods differ, their advantages and disadvantages, and when each approach is applicable.
How much does drug discovery software cost?
How number of users affect drug discovery software costs The number of people who need access to the platform is…
For QSAR models, why is the 2C9 pKi a continuous model but 2D6 is categorical?
This is due to a difference in data quality. The CYP2D6 model is categorical due to an uneven distribution in…
Do the ADME models learn from our data?
No, StarDrop’s ADME QSAR model will not learn from your data. However, if you want to build ADME models based…
How can we use our in-house data to make predictions?
All of the models in StarDrop’s ADME QSAR module are rigorously validated using external, independent test sets. You can find…
How do I find the uncertainty for each ADME QSAR model prediction?
Every model prediction is associated with an estimate of its confidence. You can see these by clicking on the ‘show…
The arrow of pKa calculations is greyed out. How can I run the pKa models?
For the pKa models, the arrow is greyed out until you select some rows in your dataset. The pKa calculations…
Do the ADME models get updated when new data is available?
The models are retrained on an ad hoc basis. The models we provide are intended to cover as much of…
Addressing toxicity risk in multi-parameter optimisation
In this example we will illustrate how knowledge-based predictions of toxicity can be used within a MPO environment to guide the selection and design of compounds with a good balance of properties and reduced risk of toxicity.
Predicting pKa using a combination of quantum mechanical and machine learning methods
This study aimed to create a model for predicting pKa using a semi-empirical quantum mechanics (QM) approach combined with machine learning (ML).
pKa prediction using quantum mechanics and machine learning
The dissociation of a proton from a heteroatom has a significant influence on the charge distribution and interactions of a…
Hit list triage
In this example, Card View is applied to analyse the results from a kinase project in which a large screening…
Guiding selection and design in hit-to-lead
The objective in this example is to identify one or more high quality chemistries for progression to detailed in vitro…
Compound prioritisation and selection
In this demo we’re going to take a look at how StarDrop can guide the prioritisation and selections of compounds using a combination of in vitro and in silico data.
