Whole-genome sequencing provides insights into the genetic diversity and domestication of bitter gourd(Momordica spp.)

Bitter gourd(Momordica charantia)is a popular cultivated vegetable in Asian and African countries.To reveal the characteristics of the genomic structure,evolutionary trajectory,and genetic basis underlying the domestication of bitter gourd,we performed whole-genome sequencing of the cultivar Dali-11 and the wild small-fruited line TR and resequencing of 187 bitter gourd germplasms from 16 countries.The major gene clusters(Bi clusters)for the biosynthesis of cucurbitane triterpenoids,which confer a bitter taste,are highly conserved in cucumber,melon,and watermelon.Comparative analysis among cucurbit genomes revealed that the Bi cluster involved in cucurbitane triterpenoid biosynthesis is absent in bitter gourd.Phylogenetic analysis revealed that the TR group,including 21 bitter gourd germplasms,may belong to a new species or subspecies independent from M.charantia.Furthermore,we found that the remaining 166 M.charantia germplasms are geographically differentiated,and we identified 710,412,and 290 candidate domestication genes in the South Asia,Southeast Asia,and China populations,respectively.This study provides new insights into bitter gourd genetic diversity and domestication and will facilitate the future genomics-enabled improvement of bitter gourd.

A high-quality sponge gourd(Luffa cylindrica)genome

Sponge gourd(Luffa cylindrica)is an important cultivated vegetable and medicinal plant in the family Cucurbitaceae.In this study,a draft genome sequence of the sponge gourd inbred line P93075 was analyzed.Using Illumina,PacBio,and 10×Genomics sequencing techniques as well as new assembly techniques such as FALCON and chromatin interaction mapping(Hi-C),a chromosome-scale genome of approximately 656.19 Mb,with an N50 scaffold length of 48.76 Mb,was generated.From this assembly,25,508 protein-coding gene loci were identified,and 63.81%of the whole-genome consisted of transposable elements,which are major contributors to the expansion of the sponge gourd genome.According to a phylogenetic analysis of conserved genes,the sponge gourd lineage diverged from the bitter gourd lineage approximately 41.6 million years ago.Additionally,many genes that respond to biotic and abiotic stresses were found to be lineage specific or expanded in the sponge gourd genome,as demonstrated by the presence of 462 NBS-LRR genes,a much greater number than are found in the genomes of other cucurbit species;these results are consistent with the high stress resistance of sponge gourd.Collectively,our study provides insights into genome evolution and serves as a valuable reference for the genetic improvement of sponge gourd.

Optimizing energy efficiency of CNN-based object detection with dynamic voltage and frequency scaling

On the one hand,accelerating convolution neural networks(CNNs)on FPGAs requires ever increasing high energy efficiency in the edge computing paradigm.On the other hand,unlike normal digital algorithms,CNNs maintain their high robustness even with limited timing errors.By taking advantage of this unique feature,we propose to use dynamic voltage and frequency scaling(DVFS)to further optimize the energy efficiency for CNNs.First,we have developed a DVFS framework on FPGAs.Second,we apply the DVFS to SkyNet,a state-of-the-art neural network targeting on object detection.Third,we analyze the impact of DVFS on CNNs in terms of performance,power,energy efficiency and accuracy.Compared to the state-of-the-art,experimental results show that we have achieved 38%improvement in energy efficiency without any loss in accuracy.Results also show that we can achieve 47%improvement in energy efficiency if we allow 0.11%relaxation in accuracy.

Smart ring resonator–based sensor for multicomponent chemical analysis via machine learning

We demonstrate a smart sensor for label-free multicomponent chemical analysis using a single label-free ring resonator to acquire the entire resonant spectrum of the mixture and a neural network model to predict the composition for multicomponent analysis. The smart sensor shows a high prediction accuracy with a low rootmean-squared error ranging only from 0.13 to 2.28 mg/m L. The predicted concentrations of each component in the testing dataset almost all fall within the 95% prediction bands. With its simple label-free detection strategy and high accuracy, the smart sensor promises great potential for multicomponent analysis applications in many fields.