High-density single nucleotide polymorphism (SNP) genotyping arrays are powerful tools that can measure the level of genetic polymorphism within a population. To develop a whole-genome SNP array for oil palms, SNP discovery was performed using deep resequencing of eight libraries derived from 132 Elaeis guineensis and Elaeis oleifera palms belonging to 59 origins, resulting in the discovery of >3 million putative SNPs. After SNP filtering, the Illumina OP200K custom array was built with 170 860 successful probes. Phenetic clustering analysis revealed that the array could distinguish between palms of different origins in a way consistent with pedigree records. Genome-wide linkage disequilibrium declined more slowly for the commercial populations (ranging from 120 kb at r(2) = 0.43 to 146 kb at r(2) = 0.50) when compared with the semi-wild populations (19.5 kb at r(2) = 0.22). Genetic fixation mapping comparing the semi-wild and commercial population identified 321 selective sweeps. A genome-wide association study (GWAS) detected a significant peak on chromosome 2 associated with the polygenic component of the shell thickness trait (based on the trait shell-to-fruit; S/F %) in tenera palms. Testing of a genomic selection model on the same trait resulted in good prediction accuracy (r = 0.65) with 42% of the S/F % variation explained. The first high-density SNP genotyping array for oil palm has been developed and shown to be robust for use in genetic studies and with potential for developing early trait prediction to shorten the oil palm breeding cycle.

Genomic selection (GS) uses genome-wide markers to select individuals with the desired overall combination of breeding traits. A total of 1,218 individuals from a commercial population of Ulu Remis x AVROS (UR x AVROS) were genotyped using the OP200K array. The traits of interest included: shell-to-fruit ratio (S/F, %), mesocarp-to-fruit ratio (M/F, %), kernel-to-fruit ratio (K/F, %), fruit per bunch (F/B, %), oil per bunch (O/B, %) and oil per palm (O/P, kg/palm/year). Genomic heritabilities of these traits were estimated to be in the range of 0.40 to 0.80. GS methods assessed were RR-BLUP, Bayes A (BA), Cπ (BC), Lasso (BL) and Ridge Regression (BRR). All methods resulted in almost equal prediction accuracy. The accuracy achieved ranged from 0.40 to 0.70, correlating with the heritability of traits. By selecting the most important markers, RR-BLUP B has the potential to outperform other methods. The marker density for certain traits can be further reduced based on the linkage disequilibrium (LD). Together with in silico breeding, GS is now being used in oil palm breeding programs to hasten parental palm selection.

Gene discovery in the Malaysian giant freshwater prawn (Macrobrachium rosenbergii) has been limited to small scale data collection, despite great interest in various research fields related to the commercial significance of this species. Next generation sequencing technologies that have been developed recently and enabled whole transcriptome sequencing (RNA-seq), have allowed generation of large scale functional genomics data sets in a shorter time than was previously possible. Using this technology, transcriptome sequencing of three tissue types: hepatopancreas, gill and muscle, has been undertaken to generate functional genomics data for M. rosenbergii at a massive scale. De novo assembly of 75-bp paired end Ilumina reads has generated 102,230 unigenes. Sequence homology search and in silico prediction have identified known and novel protein coding candidate genes (∼24%), non-coding RNA, and repetitive elements in the transcriptome. Potential markers consisting of simple sequence repeats associated with known protein coding genes have been successfully identified. Using KEGG pathway enrichment, differentially expressed genes in different tissues were systematically represented. The functions of gill and hepatopancreas in the context of neuroactive regulation, metabolism, reproduction, environmental stress and disease responses are described and support relevant experimental studies conducted previously in M. rosenbergii and other crustaceans. This large scale gene discovery represents the most extensive transcriptome data for freshwater prawn. Comparison with model organisms has paved the path to address the possible conserved biological entities shared between vertebrates and crustaceans. The functional genomics resources generated from this study provide the basis for constructing hypotheses for future molecular research in the freshwater shrimp.

BackgroundGenomic selection (GS) uses genome-wide markers as an attempt to accelerate genetic gain in breeding programs of both animals and plants. This approach is particularly useful for perennial crops such as oil palm, which have long breeding cycles, and for which the optimal method for GS is still under debate. In this study, we evaluated the effect of different marker systems and modeling methods for implementing GS in an introgressed dura family derived from a Deli dura x Nigerian dura (Deli x Nigerian) with 112 individuals. This family is an important breeding source for developing new mother palms for superior oil yield and bunch characters. The traits of interest selected for this study were fruit-to-bunch (F/B), shell-to-fruit (S/F), kernel-to-fruit (K/F), mesocarp-to-fruit (M/F), oil per palm (O/P) and oil-to-dry mesocarp (O/DM). The marker systems evaluated were simple sequence repeats (SSRs) and single nucleotide polymorphisms (SNPs). RR-BLUP, Bayesian A, B, Cπ, LASSO, Ridge Regression and two machine learning methods (SVM and Random Forest) were used to evaluate GS accuracy of the traits.ResultsThe kinship coefficient between individuals in this family ranged from 0.35 to 0.62. S/F and O/DM had the highest genomic heritability, whereas F/B and O/P had the lowest. The accuracies using 135 SSRs were low, with accuracies of the traits around 0.20. The average accuracy of machine learning methods was 0.24, as compared to 0.20 achieved by other methods. The trait with the highest mean accuracy was F/B (0.28), while the lowest were both M/F and O/P (0.18). By using whole genomic SNPs, the accuracies for all traits, especially for O/DM (0.43), S/F (0.39) and M/F (0.30) were improved. The average accuracy of machine learning methods was 0.32, compared to 0.31 achieved by other methods.ConclusionDue to high genomic resolution, the use of whole-genome SNPs improved the efficiency of GS dramatically for oil palm and is recommended for dura breeding programs. Machine learning slightly outperformed other methods, but required parameters optimization for GS implementation.Electronic supplementary materialThe online version of this article (10.1186/s12863-017-0576-5) contains supplementary material, which is available to authorized users.

Oil palm is one of the most productive oil-producing crops and can store up to 90% oil in its fruit mesocarp. Oil palm fruit is a sessile drupe consisting of a fleshy mesocarp from which palm oil is extracted. Biochemical changes in the mesocarp cell walls, polyamines, and hormones at different ripening stages of oil palm fruits were studied, and the relationship between the structural and the biochemical metabolism of oil palm fruits during ripening is discussed. Time-course analysis of the changes in expression of polyamines, hormones, and cell-wall-related genes and metabolites provided insights into the complex processes and interactions involved in fruit development. Overall, a strong reduction in auxin-responsive gene expression was observed from 18 to 22 weeks after pollination. High polyamine concentrations coincided with fruit enlargement during lipid accumulation and latter stages of maturation. The trend of abscisic acid (ABA) concentration was concordant with GA₄ but opposite to the GA₃ profile such that as ABA levels increase the resulting elevated ABA/GA₃ ratio clearly coincides with maturation. Polygalacturonase, expansin, and actin gene expressions were also observed to increase during fruit maturation. The identification of the master regulators of these coordinated processes may allow screening for oil palm variants with altered ripening profiles.

Gene expression changes that occur during mesocarp development are a major research focus in oil palm research due to the economic importance of this tissue and the relatively rapid increase in lipid content to very high levels at fruit ripeness. Here, we report the development of a transcriptome-based 105,000-probe oil palm mesocarp microarray. The expression of genes involved in fatty acid (FA) and triacylglycerol (TAG) assembly, along with the tricarboxylic acid cycle (TCA) and glycolysis pathway at 16 Weeks After Anthesis (WAA) exhibited significantly higher signals compared to those obtained from a cross-species hybridization to the Arabidopsis (p-value < 0.01), and rice (p-value < 0.01) arrays. The oil palm microarray data also showed comparable correlation of expression (r2 = 0.569, p < 0.01) throughout mesocarp development to transcriptome (RNA sequencing) data, and improved correlation over quantitative real-time PCR (qPCR) (r2 = 0.721, p < 0.01) of the same RNA samples. The results confirm the advantage of the custom microarray over commercially available arrays derived from model species. We demonstrate the utility of this custom microarray to gain a better understanding of gene expression patterns in the oil palm mesocarp that may lead to increasing future oil yield.

BackgroundThe oil yield trait of oil palm is expected to involve multiple genes, environmental influences and interactions. Many of the underlying mechanisms that contribute to oil yield are still poorly understood. In this study, we used a microarray approach to study the gene expression profiles of mesocarp tissue at different developmental stages, comparing genetically related high- and low- oil yielding palms to identify genes that contributed to the higher oil-yielding palm and might contribute to the wider genetic improvement of oil palm breeding populations.ResultsA total of 3412 (2001 annotated) gene candidates were found to be significantly differentially expressed between high- and low-yielding palms at at least one of the different stages of mesocarp development evaluated. Gene Ontologies (GO) enrichment analysis identified 28 significantly enriched GO terms, including regulation of transcription, fatty acid biosynthesis and metabolic processes. These differentially expressed genes comprise several transcription factors, such as, bHLH, Dof zinc finger proteins and MADS box proteins. Several genes involved in glycolysis, TCA, and fatty acid biosynthesis pathways were also found up-regulated in high-yielding oil palm, among them; pyruvate dehydrogenase E1 component Subunit Beta (PDH), ATP-citrate lyase, β- ketoacyl-ACP synthases I (KAS I), β- ketoacyl-ACP synthases III (KAS III) and ketoacyl-ACP reductase (KAR). Sucrose metabolism-related genes such as Invertase, Sucrose Synthase 2 and Sucrose Phosphatase 2 were found to be down-regulated in high-yielding oil palms, compared to the lower yield palms.ConclusionsOur findings indicate that a higher carbon flux (channeled through down-regulation of the Sucrose Synthase 2 pathway) was being utilized by up-regulated genes involved in glycolysis, TCA and fatty acid biosynthesis leading to enhanced oil production in the high-yielding oil palm. These findings are an important stepping stone to understand the processes that lead to production of high-yielding oil palms and have implications for breeding to maximize oil production.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-017-3855-7) contains supplementary material, which is available to authorized users.

Stomatal density is an important trait for breeding selection of drought tolerant oil palms; however, its measurement is extremely tedious. To accelerate this process, we developed an automated system. Leaf samples from 128 palms ranging from nursery (1 years old), juvenile (2–3 years old) and mature (> 10 years old) were collected to build an oil palm specific stomata detection model. Micrographs were split into tiles, then used to train a stomata object detection convolutional neural network model through transfer learning. The detection model was then tested on leaf samples acquired from three independent oil palm populations of young seedlings (A), juveniles (B) and productive adults (C). The detection accuracy, measured in precision and recall, was 98.00% and 99.50% for set A, 99.70% and 97.65% for set B, and 99.55% and 99.62% for set C, respectively. The detection model was cross-applied to another set of adult palms using stomata images taken with a different microscope and under different conditions (D), resulting in precision and recall accuracy of 99.72% and 96.88%, respectively. This indicates that the model built generalized well, in addition has high transferability. With the completion of this detection model, stomatal density measurement can be accelerated. This in turn will accelerate the breeding selection for drought tolerance.