SMILES is an industry standard with a checkered standardization history. Both primary literature descriptions and online documentation left crucial points unaddressed. These limitations resulted in OpenSMILES, the first independent attempt to codify the syntax and semantics of the SMILES language.

Many problems in business and science can be cast in terms of graph matching . Given an undirected graph, a matching is a subgraph in which every node has a degree of one. It’s often important to find a matching that contains the most possible edges. This is the maximum matching problem. Although the concept is easy to understand, algorithms for finding maximum matchings tend to be complex.

Interactive programming, in which expressions are entered and evaluated in real-time, can be a powerful tool for exploring a language and problem solving. This capability is most often associated with dynamically evaluated languages such as JavaScript and Python. Compiled languages such as Java and C++ can also be used interactively, but tooling tends to lag by many years.

Like many fields, chemistry is in the midst of a machine learning transformation. Chemistry also has some peculiarities that make getting started with machine learning a challenge. What would be helpful is a workbench that makes it possible to conduct simple but illustrative studies with minimal ceremony. DeepChem is a batteries-included suite that seeks to fill this need.

The last seven years has treated front-end developers to a banquet of browser UI frameworks. Names including React, Angular, Vue, Ember, and Backbone frequently top annual roundups and job postings. A host of lesser players also makes an appearance. What gets lost in the hype and buzzwords too often is a clear description of the main problem these tools solve.

Cycles within molecular graphs (aka “rings”) play a central role in many chemistry subdisciplines. Not surprisingly, computational chemistry and cheminformatics procedures often depend on cycle perception. But cycle perception turns out to be much more complex than first appearances suggest. Part of the problem is the sheer number of approaches.

Sometimes Web applications accept molecular input in the form of SMILES strings. For example, a form may allow a list of newline-delimited SMILES to be submitted for processing. Increasingly, users expect realtime validation during these interactions. One option would be an asynchronous HTTP call in which a server validates SMILES strings. But this requires an API to be both available and properly secured.

RDKit is a cheminformatics toolkit with bindings for Python. It’s packed with functionality, deployed within multiple open source projects, and is widely-used in machine learning applications. RDKit can also be difficult to install. This article discusses the problem and a method for using RDKit within Jupyter notebooks. Installation Options The Python Package Index (aka PyPI, aka pip) is Python’s standard package manager.

The previous article in this series described a low-level, manual approach to building Python packages from Rust code. Although effective, the procedure involves writing highly repetitive code blocks in both languages. Fortunately, such repetition isn’t necessary. This article picks up where the last one left off by showing one way to automate the generation of Python packages from Rust libraries.

Python has evolved into scientific computing’s standard orchestration language. In this role, Python sits at the top of a software stack, directing the actions of components written in low-level languages below. This architecture allows Python to deliver developer productivity while the low-level language delivers performance.

What set of features uniquely characterize a given molecule? What modes of representation should be fixed or rejected, and under what conditions? Given that machine-based molecular encodings have been in use for more than sixty years, it might seem that such questions have long since been resolved. Nevertheless, the topic casts a long shadow to this day, particularly when dataset curation or property prediction come into play.