Evaluating and clustering retrosynthesis pathways with learned strategy

With recent advances in the computer-aided synthesis planning (CASP) powered by data science and machine learning, modern CASP programs can rapidly identify thousands of potential pathways for a given target molecule. However, the lack of a holistic pathway evaluation mechanism makes it challenging to systematically prioritize strategic pathways except for using some simple heuristics. Herein, we introduce a data-driven approach to evaluate the relative strategic levels of retrosynthesis pathways using a dynamic tree-structured long short-term memory (tree-LSTM) model. We first curated a retrosynthesis pathway database, containing 238k patent-extracted pathways along with ∼55 M artificial pathways generated from an open-source CASP program, ASKCOS. The tree-LSTM model was trained to differentiate patent-extracted and artificial pathways with the same target molecule in order to learn the strategic relationship among single-step reactions within the patent-extracted pathways. The model achieved a top-1 ranking accuracy of 79.1% to recognize patent-extracted pathways. In addition, the trained tree-LSTM model learned to encode pathway-level information into a representative latent vector, which can facilitate clustering similar pathways to help illustrate strategically diverse pathways generated from CASP programs.

Tree-structured long short-term memory neural model learns to understand the retrosynthesis design strategies from patent-extracted retrosynthetic pathway data.     

Study of quasi-monodisperse In2O3 nanocrystals: synthesis and optical determination

In this communication, we report a successful synthesis of quasi-monodisperse In2O3 nanocrystals with high crystallinity in a high-temperature organic solution. The average size of nanocrystals can be tuned using a dynamic injection technique. TEM and XRD investigations indicate that each nanocrystal is a single crystal. The optical determination implies that the photoluminescence behavior of these In2O3 nanocrystals is different from that of the bulk, probably due to the combination of weak quantum-confinement-effects and the nature of high crystallinity in nanocrystals.     

Comparative study of Li, Na, and K adsorptions on graphite by using ab initio method

A comprehensive study has been conducted to compare the adsorptions of alkali metals (including Li, Na, and K) on the basal plane of graphite by using molecular orbital theory calculations. All three metal atoms prefer to be adsorbed on the "middle hollow site" above a hexagonal aromatic ring. A novel phenomenon was observed, that is, Na, instead of Li or K, is the weakest among the three types of metal atoms in adsorption. The reason is that the SOMO (single occupied molecular orbital) of the Na atom is exactly at the middle point between the HOMO and the LUMO of the graphite layer in energy level. As a result, the SOMO of Na cannot form a stable interaction with either the HOMO or the LUMO of the graphite. On the other hand, the SOMO of Li and K can form a relatively stable interaction with either the HOMO or the LUMO of graphite. Why Li has a relatively stronger adsorption than K on graphite has also been interpreted on the basis of their molecular-orbital energy levels.     

Appointment-driven queueing systems with non-punctual customers

Queueing Systems - We consider a single-server queueing system where a finite number of customers arrive over time to receive service. Arrivals are driven by appointments, with a scheduled...     

On hyper-Zagreb index conditions for hamiltonicity of graphs

Czechoslovak Mathematical Journal - During the last decade, several research groups have published results on sufficient conditions for the hamiltonicity of graphs by using some topological...     

Outcome regression-based estimation of conditional average treatment effect

Annals of the Institute of Statistical Mathematics - The research is about a systematic investigation on the following issues. First, we construct different outcome regression-based estimators for...     

The finite steps of convergence of the fast thresholding algorithms with f-feedbacks in compressed sensing

Numerical Algorithms - Iterative algorithms based on thresholding, feedback, and null space tuning (NST+HT+FB) for sparse signal recovery are exceedingly effective and efficient, particularly for...     

Monge-Ampère Equation with Bounded Periodic Data

We consider the Monge-Ampère equation det $(D^2u) = f$ in $\mathbb{R}^n,$ where $f$ is a positive bounded periodic function. We prove that $u$ must be the sum of a quadratic polynomial and a periodic function. For $f ≡ 1,$ this is the classic result by Jörgens, Calabi and Pogorelov. For $f ∈ C^α,$ this was proved by Caffarelli and the first named author.