Virtual metrology of semiconductor PVD process based on combination of tree-based ensemble model.

In order to improve the accuracy of semiconductor wafer virtual metrology, and overcome the physical metrology delay of wafer acceptance test, a virtual physical vapor deposition metrology method based on combination of tree-based ensemble models is proposed to conduct online virtual metrology on semiconductor wafer electrical parameters, and use hyperparameter optimization technique to perform model optimization and to achieve real-time alarm on process deviation. This combination of tree-based ensemble model combines Bagging, Boosting, and Stacking techniques. First, based on 4 types of base learner, Random Forest, Extra-Trees, XGBoost, and lightGBM, preliminary virtual metrology is performed on wafer PVD process, and then transforms the predict results of the 4 base learners into meta feature vector as the input of meta learner lightGBM to perform further virtual metrology. The Sequential model-based optimization algorithm is used to improve the accuracy of virtual metrology. First, the initial hyperparameter of the sequential model-based optimization is initialized by using random sampling, then the combination model is approximated by the surrogate model of tree-structured Parzen estimator, and the recommended hyperparameters is obtained by using EI (Expected Improvement), and then the optimized combination model is obtained. Finally, the superiority of the method proposed in this paper is verified by studying the results comparing to the common virtual metrology methods on the PVD process. The experiment shows the result of resistivity metrology using the combination of tree-based ensemble models in the PVD process is significantly better than LASSO regression, partial least squares regression(PLSR), support vector machine(SVR), Gaussian process regression(GPR) and artificial neural network regression(ANN).

Oligogenic Effects of 16p11.2 Copy-Number Variation on Craniofacial Development.

A copy-number variant (CNV) of 16p11.2 encompassing 30 genes is associated with developmental and psychiatric disorders, head size, and body mass. The genetic mechanisms that underlie these associations are not understood. To determine the influence of 16p11.2 genes on development, we investigated the effects of CNV on craniofacial structure in humans and model organisms. We show that deletion and duplication of 16p11.2 have "mirror" effects on specific craniofacial features that are conserved between human and rodent models of the CNV. By testing dosage effects of individual genes on the shape of the mandible in zebrafish, we identify seven genes with significant effects individually and find evidence for others when genes were tested in combination. The craniofacial phenotypes of 16p11.2 CNVs represent a model for studying the effects of genes on development, and our results suggest that the associated facial gestalts are attributable to the combined effects of multiple genes.

Paternally inherited cis-regulatory structural variants are associated with autism.

The genetic basis of autism spectrum disorder (ASD) is known to consist of contributions from de novo mutations in variant-intolerant genes. We hypothesize that rare inherited structural variants in cis-regulatory elements (CRE-SVs) of these genes also contribute to ASD. We investigated this by assessing the evidence for natural selection and transmission distortion of CRE-SVs in whole genomes of 9274 subjects from 2600 families affected by ASD. In a discovery cohort of 829 families, structural variants were depleted within promoters and untranslated regions, and paternally inherited CRE-SVs were preferentially transmitted to affected offspring and not to their unaffected siblings. The association of paternal CRE-SVs was replicated in an independent sample of 1771 families. Our results suggest that rare inherited noncoding variants predispose children to ASD, with differing contributions from each parent.

Rotavirus incidence in hospitalised Hong Kong children: 1 July 1997 to 31 March 2011.

Sentinel laboratory surveillance from one hospital and passive discharge diagnosis (Clinical Management System, CMS) data from all public Hospital Authority (HA) hospitals were used to estimate disease burden and incidence of rotavirus in hospitalised Hong Kong children over 14 rotavirus seasons (1 July 1997 to 31 March 2011). A primary diagnosis of a gastroenteritis-related disorder was noted in 9.8% of children aged below 5 years, and a primary or secondary diagnosis in 11.8%. Any CMS diagnosis of rotavirus (ICD 008.61) was initially used to derive incidence estimates of rotavirus by age group. Rotavirus was recorded as any primary or any secondary diagnosis in 1.6% of children below 5 years of age. The unadjusted incidence rates per 100,000 person-years based on any CMS diagnosis of rotavirus were: 249 (0 to <1m); 612 (1 to <2m); 1066 (2 to <6m); 1383 (6 to <11m); 959 (1 to <2y); 406 (2 to <3y); 233 (3 to <4y); 124 (4 to <5y). Overall the rotavirus incidence was 1071 in children below 2 years and 542 in children below 5 years of age, with the incidence rates trending up during the time period (p=0.001). A similar but less marked upward trend (p=0.046) was noted for the incidence of all-cause gastroenteritis. Laboratory results from a single surveillance hospital (1 July 2000 to 31 March 2011) were then linked to these CMS codes to derive adjustment factors for possible over- and under-diagnosis of rotavirus based on CMS codes alone. This analysis suggested that a CMS diagnosis of rotavirus alone likely under-reported true incidence by a factor of between 1.59 and 2.02 in children below 5 years of age. Despite the availability of rotavirus vaccines in the private sector since 2006, no reduction in the incidence of hospitalisation for either rotavirus or all-cause gastroenteritis was noted in Hong Kong children below 5 years of age over 14 rotavirus seasons (1997-2011).