Genetic programming systems typically require large computational resource investments for training-set evaluations. Downsampling these sets has proven to decrease costs and improve problem-solving success, particularly with the lexicase parent selection algorithm. We investigated its effectiveness when applied to three other common selection methods and across various program synthesis problems. Our findings show that down-sampling notably enhances all three methods, indicating its potential broad applicability. Additionally, we found informed down-sampling to be more successful than its random counterpart, particularly in selection schemes maintaining diversity like lexicase selection. We conclude that down-sampling is a promising strategy for test -based genetic programming problems, irrespective of selection scheme.

The unique reproductive structure of Euphorbia species, the cyathium, has historically been considered an inflorescence despite its visual resemblance to a single complete flower. However, multiple other models have recently been proposed, including the idea that the cyathium is a flower or has some degree of floral identity. Enabled by the petty spurge ( Euphorbia peplus L.) genome and guided by the ABCDE model of floral development, we dissected petty spurge cyathia and carried out a transcriptomic analysis of the different organs. We also constructed gene phylogenies and performed dN/dS analysis on select floral genes. The E class gene SEP3 was not upregulated in the involucre as compared with the cyathophyll, indicating that the cyathium is not a complete flower. The filiform structures show upregulation of B and E class genes and transcriptomic signatures of heterochromatin formation, consistent with the hypothesis that they are reduced flowers. E. peplus homologs for the inflorescence/floral meristem genes UFO and LFY and the B class genes AP3 and PI have highly diverged sequences relative to other Euphorbiaceae species, suggesting that these reproductive patterning genes may have evolved divergent functions. We propose a new model in which the evolution of the cyathium involves two innovations: 1. altering LFY function to influence the timing of floral meristem development to create a compact flower-like inflorescence, and 2. altering the B class genes so that no perianth is formed. Our new hypothesis includes testable predictions about protein interactions and gene sequence evolution.### Competing Interest StatementThe authors have declared no competing interest.

Genetic programming systems often use large training sets to evaluate the quality of candidate solutions for selection, which is often computationally expensive. Down-sampling training sets has long been used to decrease the computational cost of evaluation in a wide range of application domains. More specifically, recent studies have shown that both random and informed down-sampling can substantially improve problem-solving success for GP systems that use the lexicase parent selection algorithm. We test whether these down-sampling techniques can also improve problem-solving success in the context of three other commonly used selection methods, fitness-proportionate, tournament, implicit fitness sharing plus tournament selection, across six program synthesis GP problems. We verified that down-sampling can significantly improve the problem-solving success for all three of these other selection schemes, demonstrating its general efficacy. We discern that the selection pressure imposed by the selection scheme does not interact with the down-sampling method. However, we find that informed down-sampling can improve problem solving success significantly over random down-sampling when the selection scheme has a mechanism for diversity maintenance like lexicase or implicit fitness sharing. Overall, our results suggest that down-sampling should be considered more often when solving test-based problems, regardless of the selection scheme in use.

The vast majority of deaf or hard-of-hearing (DHH) children are born to hearing parents. Due to a lack of immersive exposure to their natural language - sign language - they are at severe risk of language deprivation. In response to this challenge, this paper presents a novel computer-mediated communication platform named Tabletop Interactive Play System (TIPS). It serves as a test-bed to investigate technical and ethical solutions that enable hearing parents to use American Sign Language (ASL) during face-to-face play with their with their DHH children. The TIPS platform offers a variety of user options in three key aspects: (1) ASL recommendation in alignment with hearing parents’ real-time speech; (2) ASL display through different form-factors (Augmented Reality (AR) projection, tablet, and smart glasses); and (3) autonomy support to enhance users’ sense of agency and trust in the system. In this paper, we will describe the system’s design, implementation, and preliminary evaluation results.

Genetic programming systems often use large training sets to evaluate candidate solutions, which can be computationally expensive. Down-sampling training sets has long been used to decrease the computational cost of evaluation in a wide range of application domains. Indeed, recent studies have shown that both random and informed down-sampling can substantially improve problem-solving success for GP systems that use lexicase parent selection. We use the PushGP framework to experimentally test whether these down-sampling techniques can also improve problem-solving success in the context of two other commonly used selection methods, fitness-proportionate and tournament selection, across eight GP problems (four program synthesis and four symbolic regression). We verified that down-sampling can benefit the problem-solving success of both fitness-proportionate and tournament selection. However, the number of problems wherein down-sampling improved problem-solving success varied by selection scheme, suggesting that the impact of down-sampling depends both on the problem and choice of selection scheme. Surprisingly, we found that down-sampling was most consistently beneficial when combined with lexicase selection as compared to tournament and fitness-proportionate selection. Overall, our results suggest that down-sampling should be considered more often when solving test-based GP problems.

Since most deaf or hard-of-hearing (DHH) children are born into hearing families, they are at risk of lifelong language deficits from a lack of natural language exposure in early childhood. The Tabletop Interactive Play System (TIPS) is a realtime American Sign Language (ASL) communication aid, for face-to-face joint play between hearing parents and their DHH children. We investigate user preferences for different ASL recommendation methods, providing insights into optimizing ASL communication technologies for hearing parent-DHH child interactions. We introduce Tap-to-Sign, an AR projection-based ASL sign retrieval system that grants parents complete autonomy in selecting the word to sign to their DHH child. Additionally, we propose three distinct strategies to automatically recommend a word for the parent to sign. To gain insights into our diverse methods, preliminary feedback was gathered from key stakeholders.

Genetic programming systems often use large training sets to evaluate the quality of candidate solutions for selection. However, evaluating populations on large training sets can be computationally expensive. Down-sampling training sets has long been used to decrease the computational cost of evaluation in a wide range of application domains. Indeed, recent studies have shown that both random and informed down-sampling can substantially improve problem-solving success for GP systems that use the lexicase parent selection algorithm. We use the PushGP framework to experimentally test whether these down-sampling techniques can also improve problem-solving success in the context of two other commonly used selection methods, fitness-proportionate and tournament selection, across eight GP problems (four program synthesis and four symbolic regression). We verified that down-sampling can benefit the problem-solving success of both fitness-proportionate and tournament selection. However, the number of problems wherein down-sampling improved problem-solving success varied by selection scheme, suggesting that the impact of down-sampling depends both on the problem and choice of selection scheme. Surprisingly, we found that down-sampling was most consistently beneficial when combined with lexicase selection as compared to tournament and fitness-proportionate selection. Overall, our results suggest that down-sampling should be considered more often when solving test-based GP problems.

Euphorbia peplus (petty spurge) is a small, fast-growing plant that is native to Eurasia and has become a naturalized weed in North America and Australia. Euphorbia peplus is not only medicinally valuable, serving as a source for the skin cancer drug ingenol mebutate, but also has great potential as a model for latex production owing to its small size, ease of manipulation in the laboratory, and rapid reproductive cycle. To help establish E. peplus as a new model, we generated a 267.2-Mb Hi-C-anchored PacBio HiFi nuclear genome assembly with a BUSCO score of 98.5%, a genome annotation based on RNA-seq data from six organs, and publicly accessible tools including a genome browser and an interactive organ-specific expression atlas. Chromosome number is highly variable across Euphorbia species. Using a comparative analysis of our newly sequenced E. peplus genome with other Euphorbiaceae genomes, we show that variation in Euphorbia chromosome number between E. peplus and Euphorbia lathyris is likely due to fragmentation and rearrangement rather than chromosomal duplication followed by diploidization of the duplicated sequence. Moreover, we found that the E. peplus genome is relatively compact compared with related members of the genus in part due to restricted expansion of the Ty3 transposon family. Finally, we identify a large gene cluster that contains many previously identified enzymes in the putative ingenol mebutate biosynthesis pathway, along with additional gene candidates for this biosynthetic pathway. The genomic resources we have created for E. peplus will help advance research on latex production and ingenol mebutate biosynthesis in the commercially important Euphorbiaceae family.