Identification of stable quantitative trait loci (QTLs) across different environments and mapping populations is a prerequisite for marker-assisted selection (MAS) for cotton yield and fiber quality. To construct a genetic linkage map and to identify QTLs for fiber quality and yield traits, a backcross inbred line (BIL) population of 146 lines was developed from a cross between Upland cotton (Gossypium hirsutum) and Egyptian cotton (Gossypium barbadense) through two generations of backcrossing using Upland cotton as the recurrent parent followed by four generations of self pollination. The BIL population together with its two parents was tested in five environments representing three major cotton production regions in China. The genetic map spanned a total genetic distance of 2,895 cM and contained 392 polymorphic SSR loci with an average genetic distance of 7.4 cM per marker. A total of 67 QTLs including 28 for fiber quality and 39 for yield and its components were detected on 23 chromosomes, each of which explained 6.65-25.27% of the phenotypic variation. Twenty-nine QTLs were located on the At subgenome originated from a cultivated diploid cotton, while 38 were on the Dt subgenome from an ancestor that does not produce spinnable fibers. Of the eight common QTLs (12%) detected in more than two environments, two were for fiber quality traits including one for fiber strength and one for uniformity, and six for yield and its components including three for lint yield, one for seedcotton yield, one for lint percentage and one for boll weight. QTL clusters for the same traits or different traits were also identified. This research represents one of the first reports using a permanent advanced backcross inbred population of an interspecific hybrid population to identify QTLs for fiber quality and yield traits in cotton across diverse environments. It provides useful information for transferring desirable genes from G. barbadense to G. hirsutum using MAS.

The ‘You Only Look Once’ v3 (YOLOv3) method is among the most widely used deep learning-based object detection methods. It uses the k-means cluster method to estimate the initial width and height of the predicted bounding boxes. With this method, the estimated width and height are sensitive to the initial cluster centers, and the processing of large-scale datasets is time-consuming. In order to address these problems, a new cluster method for estimating the initial width and height of the predicted bounding boxes has been developed. Firstly, it randomly selects a couple of width and height values as one initial cluster center separate from the width and height of the ground truth boxes. Secondly, it constructs Markov chains based on the selected initial cluster and uses the final points of every Markov chain as the other initial centers. In the construction of Markov chains, the intersection-over-union method is used to compute the distance between the selected initial clusters and each candidate point, instead of the square root method. Finally, this method can be used to continually update the cluster center with each new set of width and height values, which are only a part of the data selected from the datasets. Our simulation results show that the new method has faster convergence speed for initializing the width and height of the predicted bounding boxes and that it can select more representative initial widths and heights of the predicted bounding boxes. Our proposed method achieves better performance than the YOLOv3 method in terms of recall, mean average precision, and F1-score.

No abstract

No abstract