A Data-Driven Approach to Predicting Tablet Properties after Accelerated Test Using Raw Material Property Database and Machine Learning

The purpose of this study was to develop a model for predicting tablet properties after an accelerated test and to determine whether molecular descriptors affect tablet properties. Tablets were prepared using 81 types of active pharmaceutical ingredients, with the same formulation and three different levels of compression pressure. The tablet properties measured were the tensile strength and disintegration time of tablets after two weeks of accelerated test. The material properties measured were the change in tablet thickness before and after the accelerated test, maximum swelling force, swelling time, and swelling rate. The acquired data were added to our previously constructed database containing a total of 20 material properties and 3381 molecular descriptors. The feature importance values of molecular descriptors, material properties and the compression pressure for each tablet property were calculated by random forest, which is one type of machine learning (ML) that uses ensemble learning and decision trees. The results showed that more than half of the top 25 most important features were molecular descriptors for both tablet properties, indicating that molecular descriptors are strongly related to tablet properties. A prediction model of tablet properties was constructed by eight ML types using 25 of the most important features. The results showed that the boosted neural network exhibited the best prediction accuracy and was able to predict tablet properties with high accuracy. A data-driven approach is useful for discovering intricate relationships hidden within complex and large data sets and predicting tablet properties after an accelerated test.

Application of machine learning to a material library for modeling of relationships between material properties and tablet properties

This study investigates the usefulness of machine learning for modeling complex relationships in a material library. We tested 81 types of active pharmaceutical ingredients (APIs) and their tablets to construct the library, which included the following variables: 20 types of API material properties, one type of process parameter (three levels of compression pressure), and two types of tablet properties (tensile strength (TS) and disintegration time (DT)). The machine learning algorithms boosted tree (BT) and random forest (RF) were applied to analysis of our material library to model the relationships between input variables (material properties and compression pressure) and output variables (TS and DT). The calculated BT and RF models achieved higher performance statistics compared with a conventional modeling method (i.e., partial least squares regression), and revealed the material properties that strongly influence TS and DT. For TS, true density, the tenth percentile of the cumulative percentage size distribution, loss on drying, and compression pressure were of high relative importance. For DT, total surface energy, water absorption rate, polar surface energy, and hygroscopicity had significant effects. Thus, we demonstrate that BT and RF can be used to model complex relationships and clarify important material properties in a material library.

A method for forensic identification of vegetable oil stains--rapid analysis of carboxylic acids with methyl esterification using purge-and-trap gas chromatography/mass spectrometry

A simple method using purge-and-trap gas chromatography/mass spectrometry (P&T-GC/MS) for forensic examination of oil stains was studied. Carboxylic acids, chosen as target components for discrimination of oil samples, were extracted from stains with ether, methyl esterified by tetramethylammonium hydroxide, and analyzed by P&T-GC/MS. Vegetable oils were discriminated according to their carboxylic acid compositions. Carboxylic acid composition was independent of the substrate material of the stain. Although the carboxylic acid composition of the oil changed on exposure to sunlight, identification of oil was possible for oil stains that had been in the shade, if analysis was made within 20 days.

Screening of soil evidence by a combination of simple techniques: validity of particle size distribution

The validity of the determination of particle size distribution for forensic soil identification was studied using 73 soil samples collected from a 20 km x 15 km area. Soil particles were first divided into two fractions by wet sieving with a sieve with a 0.05 mm aperture size. Dry sieving was used to analyze the coarse particle fraction and a particle size analyzer was used for the fine particle fraction. The number of particle size classes measured by sieving could be reduced to three, <0.05, 0.05-0.2, and 0.2-2 mm, which enabled 87.9% of the soil samples to be discriminated. Measurements made with the particle size analyzer on the fine particle fraction increased the discriminating power to 95.9%. The use of particle size analysis by sieving in combination with color examination allowed us to discriminate 99.5% of the soil samples, even when particle size analysis of the fine particle fraction was not carried out.

Drunken driving and breath alcohol test at the scene of violence in Japan

Road Traffic Law prescribes that no person shall drive any vehicle under the influence of intoxicating liquor. Actually, determination of "influence of alcohol" is based on the standard set by the Cabinet Order that alcohol level exceeds 0.5 mg/ml of blood or 0.25 mg/l of expiration. In 1988, number of traffic accidents by drunken driving is 4,808 cases. Among the accidents by drunken driving the rate of fatal causes is 12%. During about last ten years, both of accidents and fatal cases by drunken driving have been decreasing in number, on the other hand, charged or cited number of violation involved "driving under the influence of alcohol" has been increasing. This fact indicates that the concept of seriousness of driving after heavy drinking has been diffused over Japanese nation, but there is still a tendency among drivers to consider the effect of alcohol on driving operation to be negligible when light drinking. In a sobriety checkpoint, alcohol field test are carried out on breath to screen out impaired driver. The most common device used in Japan is an alcohol detector tube, in which cerite particles coated with chromate are packed. The scale on the detector tube is marked to show a value that is lower by 20% than the actual value concerning its inaccuracy. In our study on accuracy of the alcohol detector tube using samples containing approximately 0.25 mg/l of alcohol, which is legal critical level of alcohol impairment, coefficients of variation were 1.50 to 5.45% and deviations from the analytical results by gas chromatography were 18.2 to 19.5%.

[Craniomandibular disorders management (1) Treatment planning]

Various treatment modalities have been reported in the management of the craniomandibular disorders in the half century after Costen reported his hypothesis on the temporomandibular disorders. Recent advances in the basic sciences such as anatomy, histology, pathology, physiology and clinical trials have brought us to this new era. We have been trying to establish our diagnosis and treatment modalities in the management of the craniomandibular disorders over the past fifteen years. In this series of articles we will describe our present concept. We will first discuss diagnosis and treatment planning. After the diagnosis and initial treatment planning craniomandibular management is usually divided into two phases. Phase I is the treatment of the stomatognathic system especially the temporomandibular joint. Behavioral modification and physical therapy and splint therapy are the most common forms of treatment. Sometimes, surgical treatment is required. Phase II is the compensatory phase of the occlusion consisting of occlusal equilibration, occlusal restorations, orthodontic treatment and gnathic surgery if needed. Another phase we like to add is maintenance. In this phase, home therapy, corrective follow-up and long term observation should be emphasized. The motivation of the patient throughout the management of the craniomandibular disorder (phase I, phase II and the maintenance phase) is essential to reach to a successful result.

Purification and mode of action of a microsomal endoribonuclease from rat liver

An endoribonuclease has been purified nearly to homogeneity from rat liver microsomes, and its mode of action and general properties were studied. The enzyme had an apparent molecular weight of 58 000, as estimated by both gel filtration and SDS-polyacrylamide gel electrophoresis and produced oligonucleotides from poly(A), poly(U) and poly(C). No mononucleotide was obtained by the enzymatic hydrolysis of the above substrates. The enzyme made endonucleolytic cleavages which generated 5'-phosphate-terminated oligonucleotides. It was suggested that the existence of at least (Ado5'P)2 residues at both sides of the cleavage bond was necessary for the action of the endoribonuclease. Divalent cations (Mg2+ or Mn2+) were required for the enzymatic activity, while K+ inhibited the enzyme. Spermine stimulated the enzymatic activity in the presence of 1 mM Mg2+.