Algebraic graph-assisted bidirectional transformers for molecular property prediction

An Algebraic Property Clustering Technique for Molecular Design

The introduction of the property integration framework has enabled representation of processes and products from a properties perspective. The framework enables identification of the desired component properties by targeting optimum process performance without committing to any components during the solution step. To provide a unifying methodology for handling both process and molecular design ...

Graph Memory Networks for Molecular Activity Prediction

Molecular activity prediction is critical in drug design. Machine learning techniques such as kernel methods and random forests have been successful for this task. These models require fixed-size feature vectors as input while the molecules are variable in size and structure. As a result, fixed-size fingerprint representation is poor in handling substructures for large molecules. Here we approa...

PLDI: U: Type Assisted Synthesis of Recursive Transformers on Algebraic Datatypes

As programming languages are being developed to be used in a wide range of industrial applications, it is necessary to introduce convenience language constructs that abstract a lot of the underlying low-level code. For example, the for statement in languages like Java and C is a convenience construct that can be purely expressed in terms of the while statement. Compilers for these languages, ho...

An Algebraic Construction of Predicate Transformers

In this paper we present an algebraic construction of monotonic predicate transformers, using a categorical construction which is similar to the algebraic construction of the integers from the natural numbers. When applied to the category of sets and total functions once, it yields a category isomorphic to the category of sets and relations; a second application yields a category isomorphic to ...

Selection of Molecular Descriptor Subsets for Property Prediction