Genomic prediction for testes weight of the tiger pufferfish, Takifugu rubripes, using medium to low density SNPs

Aquaculture production is expected to increase with the help of genomic selection (GS). The possibility of performing GS using only a small number of SNPs has been examined in order to reduce genotyping costs; however, the practicality of this approach is still unclear. Here, we tested whether the effects of reducing the number of SNPs impaired the prediction accuracy of GS for standard length, body weight, and testes weight in the tiger pufferfish (Takifugu rubripes). High values for predictive ability (0.563–0.606) were obtained with 4000 SNPs for all traits under a genomic best linear unbiased predictor (GBLUP) model. These values were still within an acceptable range with 1200 SNPs (0.554–0.588). However, predictive abilities and prediction accuracies deteriorated using less than 1200 SNPs largely due to the reduced power in accurately estimating the genetic relationship among individuals; family structure could still be resolved with as few as 400 SNPs. This suggests that the SNPs informative for estimation of genetic relatedness among individuals differ from those for inference of family structure, and that non-random SNP selection based on the effects on family structure (e.g., site-FST, principal components, or random forest) is unlikely to increase the prediction accuracy for these traits.

Genetic analysis of three maize husk traits by QTL mapping in a maize-teosinte population

Background The maize husk consists of numerous leafy layers and plays vital roles in protecting the ear from pathogen infection and dehydration. Teosinte, the wild ancestor of maize, has about three layers of small husk outer covering the ear. Although several quantitative trait loci (QTL) underlying husk morphology variation have been reported, the genetic basis of husk traits between teosinte and maize remains unclear. Results A linkage population including 191 BC2F8 inbred lines generated from the maize line Mo17 and the teosinte line X26–4 was used to identify QTL associated with three husk traits: i.e., husk length (HL), husk width (HW) and the number of husk layers (HN). The best linear unbiased predictor (BLUP) depicted wide phenotypic variation and high heritability of all three traits. The HL exhibited greater correlation with HW than HN. A total of 4 QTLs were identified including 1, 1, 2, which are associated with HL, HW and HN, respectively. The proportion of phenotypic variation explained by these QTLs was 9.6, 8.9 and 8.1% for HL, HN and HW, respectively. Conclusions The QTLs identified in this study will pave a path to explore candidate genes regulating husk growth and development, and benefit the molecular breeding program based on molecular marker-assisted selection to cultivate maize varieties with an ideal husk morphology.

Simulation Study of Robust Geographically Weighted Empirical Best Linear Unbiased Predictor on Small Area Estimation

Small area estimation can be used to predict the population parameter with small sample sizes. For some cases, the population units that are close spatially may be more related than units that are further apart. The use of spatial information like geographic coordinates are studied in this research. Outlier contaminations can affect small area estimations. This study was conducted using simulation methods on generated data with six scenarios. The scenarios are the combination of spatial effects (spatial stationary and spatial non-stationary) with outlier contamination (no outlier, symmetric outliers, and non-symmetric outliers). The purpose of this study was to compare the geographically weighted empirical best linear unbiased predictor (GWEBLUP) and robust GWEBLUP (RGWEBLUP) with direct estimator, EBLUP, and REBLUP using simulation data. The performance of the predictors is evaluated using relative root mean squared error (RRMSE). The simulation results showed that geographically weighted predictors have the smallest RRMSE values for scenarios with spatial non-stationary, therefore offer a better prediction. For scenarios with outliers, robust predictors with smaller RRMSE values offer more efficiency than non-robust predictors.

PENDUGAAN AREA KECIL UNTUK ANGKA PARTISIPASI KASAR PENDIDIKAN DASAR DAN MENENGAH TINGKAT KABUPATEN/KOTA DI PROVINSI JAWA TENGAH TAHUN 2018

Pada Maret tahun 2018, Angka Partisipasi Kasar (APK) di Indonesia untuk pendidikan dasar dan menengah adalah sebesar: APK SD/sederajat 108,61%, APK SMP/sederajat 91,52%, sedangkan APK SMA/sederajat 80,68%. Capaian tersebut masih jauh dari target Rencana Pembangunan Jangka Menengah Nasional (RPJMN) 2014-2019. Salah satu provinsi yang memiliki APK pendidikan dasar dan menengah di bawah target RPJMN adalah Provinsi Jawa Tengah. Upaya yang dapat dilakukan untuk mewujudkan target tersebut adalah dengan mengetahui capaian APK pendidikan dasar dan menengah di level kabupaten/kota berdasarkan hasil Susenas September sehingga kontrol dapat dilakukan dua kali dalam setahun. Namun, langkah ini akan memerlukan penambahan jumlah sampel yang menyebabkan diperlukannya waktu, biaya, tenaga dan pemikiran yang lebih besar. Untuk mengatasi hal tersebut, Small Area Estimation (SAE) dapat digunakan untuk menghasilkan presisi yang memadai tanpa melakukan penambahan jumlah sampel. SAE merupakan metode pendugaan parameter-parameter subpopulasi yang memiliki ukuran sampel kecil. Metode SAE yang banyak digunakan adalah Empirical Best Linear Unbiased Predictor (EBLUP). Namun, model ini belum memasukkan pengaruh spasial ke dalam model. Model Fay-Herriot yang memerhatikan efek spasial dikenal dengan Spatial Empirical Best Linear Unbiased Predictor (SEBLUP). Hasil penelitian menunjukkan bahwa metode EBLUP lebih baik dalam mengestimasi APK SD/sederajat dan APK SMA/sederajat, dan metode SEBLUP lebih baik dalam mengestimasi APK SMP/sederajat.

Artificial Neural Networks in the Prediction of Genetic Merit to Flowering Traits in Bean Cultivars

Flowering is an important agronomic trait that presents non-additive gene action. Genome-enabled prediction allow incorporating molecular information into the prediction of individual genetic merit. Artificial neural networks (ANN) recognize patterns of data and represent an alternative as a universal approximation of complex functions. In a Genomic Selection (GS) context, the ANN allows automatically to capture complicated factors such as epistasis and dominance. The objectives of this study were to predict the individual genetic merits of the traits associated with the flowering time in the common bean using the ANN approach, and to compare the predictive abilities obtained for ANN and Ridge Regression Best Linear Unbiased Predictor (RR-BLUP). We used a set of 80 bean cultivars and genotyping was performed with a set of 384 SNPs. The higher accuracy of the selective process of phenotypic values based on ANN output values resulted in a greater efficacy of the genomic estimated breeding value (GEBV). Through the root mean square error computational intelligence approaches via ANN, GEBV were shown to have greater efficacy than GS via RR-BLUP.

Ensemble Learning of Convolutional Neural Network, Support Vector Machine, and Best Linear Unbiased Predictor for Brain Age Prediction: ARAMIS Contribution to the Predictive Analytics Competition 2019 Challenge

We ranked third in the Predictive Analytics Competition (PAC) 2019 challenge by achieving a mean absolute error (MAE) of 3.33 years in predicting age from T1-weighted MRI brain images. Our approach combined seven algorithms that allow generating predictions when the number of features exceeds the number of observations, in particular, two versions of best linear unbiased predictor (BLUP), support vector machine (SVM), two shallow convolutional neural networks (CNNs), and the famous ResNet and Inception V1. Ensemble learning was derived from estimating weights via linear regression in a hold-out subset of the training sample. We further evaluated and identified factors that could influence prediction accuracy: choice of algorithm, ensemble learning, and features used as input/MRI image processing. Our prediction error was correlated with age, and absolute error was greater for older participants, suggesting to increase the training sample for this subgroup. Our results may be used to guide researchers to build age predictors on healthy individuals, which can be used in research and in the clinics as non-specific predictors of disease status.

Genomic selection for heterobothriosis resistance concurrent with body size in the tiger pufferfish, Takifugu rubripes

Parasite resistance traits in aquaculture species often have moderate heritability, indicating the potential for genetic improvements by selective breeding. However, parasite resistance is often synonymous with an undesirable negative correlation with body size. In this study, we first tested the feasibility of genomic selection (GS) on resistance to heterobothriosis, caused by the monogenean parasite Heterobothrium okamotoi, which leads to huge economic losses in aquaculture of the tiger pufferfish Takifugu rubripes. Then, using a simulation study, we tested the possibility of simultaneous improvement of parasite resistance, assessed by parasite counts on host fish (HC), and standard length (SL). Each trait showed moderate heritability (square-root transformed HC: h2 = 0.308 ± 0.123, S.E.; SL: h2 = 0.405 ± 0.131). The predictive abilities of genomic prediction among 12 models, including genomic Best Linear Unbiased Predictor (GBLUP), Bayesian regressions, and machine learning procedures, were also moderate for both transformed HC (0.248‒0.344) and SL (0.340‒0.481). These results confirmed the feasibility of GS for this trait. Although an undesirable genetic correlation was suggested between transformed HC and SL (rg = 0.228), the simulation study suggested the desired gains index can help achieve simultaneous genetic improvements in both traits.