Biology subjects

You, N.

Publications and source records attributed to You, N..

4 recordsLinked to original sources

bioRxiv2026

ABC1K7: A 350-Myr chloroplast rheostat fine-tuned during coconut domestication

Perennial crops follow different domestication trajectories from annuals, yet the molecular basis of slow-variable domestication--subtle tuning of conserved regulatory hubs--remains poorly characterized. We reconstructed the evolutionary history of the chloroplast kinase ABC1K7 across 9 seed plant species spanning [~]350 Myr, employing PAML codon models, IQ-TREE robust codon models, and protein-level phylogenetic inference, with AlphaFold2 structural modeling. ABC1K7 was under extreme purifying selection ({omega} = 0.073-0.104) across all seed plants. In coconut, a single Y[->]F substitution at residue 652--located >30 [A] from the catalytic core in a predicted intrinsically disordered region--represents the only non-synonymous change differentiating coconut from 7 of 8 angiosperm orthologs, and exhibits perfect co-segregation with domestication traits across a 17-year breeding panel (n = 327). These findings provide population-level evidence consistent with the slow-variable domestication model, identifying ABC1K orthologs as targets for perennial crop improvement.

genomics↗

bioRxiv2026

A missing PEPC1 exaptation restricts C₄ evolution in palms (Arecaceae)

O_LIC4 and CAM photosynthesis evolved repeatedly across angiosperms, yet the four palm species examined here -- representing three of five Arecaceae subfamilies ([~]2,600 species total) -- all lack both. We hypothesized that this reflects the absence of PEPC1, a phosphoenolpyruvate carboxylase isoform required for C4 carbon fixation initiation and present in commelinids (Poaceae + Bromeliaceae) but whose distribution across monocots is unknown. C_LIO_LIWe surveyed six core C4 enzyme families by HMMER profiling across nine plant genomes -- four palms (Cocos nucifera, Elaeis guineensis, Phoenix dactylifera, Nypa fruticans), two grasses, one bromeliad, one basal monocot, and one fern. We constructed maximum-likelihood PEPC phylogenies and performed codon-based branch-site likelihood ratio tests using PAML. C_LIO_LIAll six enzyme families were present in palms. PEPC1 was detected in all commelinids (rice, maize, pineapple) but absent from all four palm species and outgroups. No palm PEPC sequence fell within the PPC-1 clade. Molecular dating indicates the Arecaceae-commelinid divergence ([~]120 Mya) predates PEPC1 origin ([~]105 Mya). Branch-site tests on the maize C4-PEPC1 branch were marginally significant (2{Delta}l = 3.03, p {approx} 0.082). C_LIO_LIThe absence of PEPC1 represents an ancestral molecular barrier that likely precluded C4 evolution in palms. This identifies a lineage-specific gene duplication as a constraint on photosynthetic pathway evolution in a major tropical plant lineage. C_LI

plant biology↗

bioRxiv2026

PEPC Gene Family Evolution Across Arecaceae: Complete Five-Subfamily Evidence for a Two-Lock Model of Irreversible C4 Exclusion

BackgroundC4 photosynthesis has evolved more than sixty times across flowering plants, but never in palms. Over 2,500 palm species have spent more than 100 million years in high-light, water-stressed habitats where C4 physiology would be advantageous, yet none have taken this path. Phosphoenolpyruvate carboxylase (PEPC) is the enzyme that gates entry into the C4 pathway. In every C4 grass, a commelinid-specific paralog called PEPC1 drives the CO2-concentrating mechanism that defines the syndrome. Until this study, one of the five palm subfamilies -- Ceroxyloideae -- had never been examined for PEPC gene content. ResultsWe present a complete PEPC census across 13 palm species spanning all five subfamilies (Calamoideae, Nypoideae, Coryphoideae, Ceroxyloideae, Arecoideae). Ceroxyloideae was characterized for the first time, using deep tBLASTn screening of ten million whole-genome shotgun reads. PEPC1 is absent from every palm species examined. Copy numbers are tightly constrained to 5-7 per genome, in stark contrast with the 6-25 copies found in grasses. A phylogenetic analysis of 337 PEPC sequences places all 55 palm loci at basal positions relative to the PEPC1 clade, consistent with the palm lineage having diverged approximately 120 million years ago -- roughly 15 million years before the PEPC1 duplication that occurred within the commelinids. Birth-death modeling reveals long-term gene family stasis in palms. Codon substitution analyses yield a genome-wide {omega} of 0.0582, indicating pervasive purifying selection and no signature of C4-specific adaptation. Coconut PEPC promoters uniformly lack the mesophyll-specific MESP-1/GT-1/DOF regulatory module that is required for C4-type gene expression. ConclusionsWe propose a Two-Lock Model of irreversible C4 exclusion in palms. The first is a Gene Lock: PEPC1 is phylogenetically absent from all five subfamilies. The second is a Regulatory Lock: the ancestral promoter architecture of palm PEPC genes lacks the cis-elements needed for mesophyll-specific expression. Together, these two locks -- one phylogenetic, one regulatory -- explain why C4 photosynthesis has never arisen in one of the largest C4-absent plant families. The model provides a testable framework for investigating other lineages that have remained closed to the C4 path.

evolutionary biology↗

bioRxiv2025

Benchmarking Pre-trained Genomic Language Models for RNA Sequence-Related Predictive Applications

RNA plays a pivotal role in diverse cellular functions across organisms. Developing computational algorithms for RNA sequence related questions is highly valuable. Recently, genomic language models (gLMs) with pre-training have emerged, offering flexibility for various downstream prediction tasks. However, comprehensive and fair evaluations of gLMs are lacking. In this study, we benchmark eight gLMs on prediction tasks covering four RNA processes, highlighting their strengths and limitations. While gLMs excel in performance overall, the larger model is not always better. Interestingly, models that integrate biological information consistently perform well in related tasks. Notably, gLMs demonstrate superior performance with limited training data, whereas task-specific methods achieve comparable performance with better computational efficiency when sufficient training data is available. Finally, we provide recommendations for model selection in different scenarios. These evaluation results underscore the potential of gLMs and suggest areas for future improvement.

genomics↗