Biology subjects

Alvarez, M. A.

Publications and source records attributed to Alvarez, M. A..

4 recordsLinked to original sources

bioRxiv2026

NexuST: A Hierarchical Foundation Model for Spatial Transcriptomics

Spatial transcriptomics captures molecular states within cells and their organisation in tissue. However, integrating fine-grained gene information with spatial context at scale remains challenging for existing foundation models. Here we present NexuST, a hierarchical foundation model that repeatedly interleaves gene-level molecular modelling with cell-level spatial modelling, allowing the two levels to refine one another during end-to-end pretraining. For pretraining, we curated HumanST-46M, comprising 45.7 million human cells from 72 datasets across 11 organs and three imaging-based platforms. Across four held-out datasets totalling approximately 2.6 million cells, NexuST achieved state-of-the-art or competitive performance in cell-type annotation, region prediction, gene recovery and neighbourhood-composition prediction. We find that cell-intrinsic expression remains informative even for spatial tasks, as shown by an expression-only PCA baseline, while NexuST shows particularly strong gains where spatial context is essential. Overall, NexuST establishes a hierarchical framework that can serve as a general backbone for future spatial transcriptomics foundation models.

bioinformatics↗

bioRxiv2026

Proteomics-constrained deconvolution reveals spatial cell-type programs in tumours

Accurately resolving cell-type mixtures in spatial transcriptomics remains challenging, particularly in heterogeneous tumours where cell populations are intermixed and matched single-cell references may be unavailable or poorly aligned. Current deconvolution approaches either require high-quality scRNA-seq references, suffer from scalability limitations, or lack interpretability. We introduce PISTACHIO, a proteomics-informed spatial transcriptomics deconvolution framework based on constrained non-negative matrix factorization with a negative-binomial likelihood. Rather than using probabilistic priors, PISTACHIO incorporates spatial cell-type constraints derived from paired Imaging Mass Cytometry, enforcing biologically grounded sparsity and explicit spatial feasibility of cell-type presence. PISTACHIO improved recovery of spatial cell-type distributions compared with Cell2location and STdeconvolve across synthetic and real tumour datasets. Our approach remains robust under cell-type assignment errors, maintaining high correlation with ground-truth under moderate noise, and achieves fast runtime on standard hardware, enabling practical large-scale deployment.

bioinformatics↗

bioRxiv2022

EARLY FLOWERING 3 interactions with PHYTOCHROME B and PHOTOPERIOD1 are critical for the photoperiodic regulation of wheat heading time

The photoperiodic response is critical for plants to adjust their reproductive phase to the most favorable season. Wheat heads earlier under long days (LD) than under short days (SD) and this difference is mainly regulated by the PHOTOPERIOD1 (PPD1) gene. Tetraploid wheat plants carrying the Ppd-A1a allele with a large deletion in the promoter head earlier under SD than plants carrying the wildtype Ppd-A1b allele with an intact promoter. Phytochromes PHYB and PHYC are necessary for the light activation of PPD1, and mutations in either of these genes result in the downregulation of PPD1 and very late heading time. We show here that both effects are reverted when the phyB mutant is combined with loss-of-function mutations in EARLY FLOWERING 3 (ELF3), a component of the Evening Complex (EC) in the circadian clock. We also show that the wheat ELF3 protein interacts with PHYB and PHYC, is rapidly modified by light, and binds to the PPD1 promoter in planta (likely as part of the EC). Deletion of the ELF3 binding region in the Ppd-A1a promoter results in its upregulation at dawn, similar to PPD1 alleles with intact promoters in the elf3 mutant background. The upregulation of PPD1 is correlated with the upregulation of the florigen gene FLOWERING LOCUS T1 (FT1) and early heading time. Loss-of-function mutations in ppd1 result in the downregulation of FT1 and delayed heading, even when combined with the elf3 mutation. Taken together, these results indicate that ELF3 operates downstream of PHYB as a direct transcriptional repressor of PPD1, and that this repression is relaxed both by light and by the deletion of the ELF3 binding region in the Ppd-A1a promoter. In summary, the regulation of the light mediated activation of PPD1 by ELF3 is critical for the photoperiodic regulation of wheat heading time. Author SummaryThe coordination of reproductive development with the optimal season for seed production is critical to maximize grain yield in crop species. Plants can perceive the length of the day or night (photoperiod) and use this information to anticipate seasonal changes. In most eudicot plants, CONSTANS plays a central role in the perception of photoperiod, but in wheat the main photoperiod gene is PHOROPERIOD1 (PPD1). In this study, we show that the clock gene EARLY FLOWERING 3 (ELF3) regulates the phytochrome-mediated light activation of PPD1. Loss-of-function mutations in ELF3 result in the upregulation of PPD1 at and dawn, and in early heading under both long and short days, even in the absence of PHYB. A deletion in the PPD1 promoter including an ELF3 binding region also results in earlier heading under short days, indicating that ELF3 acts as a direct transcriptional repressor of PPD1. This study shows that ELF3 plays a critical role in the wheat photoperiod pathway by regulating the light signal between the phytochromes and PPD1. ELF3 provides an additional entry point to engineer heading time in wheat, an important trait for the development of better adapted varieties to a changing environment.

plant biology↗

bioRxiv2020

Epistatic interactions between PHOTOPERIOD-1, CONSTANS 1 and CONSTANS 2 modulate the photoperiodic response in wheat

In Arabidopsis, CONSTANS (CO) integrates light and circadian clock signals to promote flowering under long days (LD). In the grasses, a duplication generated two paralogs designated as CONSTANS 1 (CO1) and CONSTANS 2 (CO2). Here we show that in tetraploid wheat plants grown under LD, combined loss-of-function mutations in the A and B-genome homeologs of CO1 and CO2 (co1 co2) result in a small (3 d) but significant (P<0.0001) acceleration of heading time both in PHOTOPERIOD1 (PPD1) sensitive (Ppd-A1b, functional ancestral allele) and insensitive (Ppd-A1a, functional dominant allele) backgrounds. Under short days (SD), co1 co2 mutants headed 13 d earlier than the wild type (P<0.0001) in the presence of Ppd-A1a. However, in the presence of Ppd-A1b, spikes from both genotypes failed to emerge by 180 d. These results indicate that CO1 and CO2 operate mainly as weak heading time repressors in both LD and SD. By contrast, in ppd1 mutants with loss-of-function mutations in both PPD1 homeologs, the wild type Co1 allele accelerated heading time >60 d relative to the co1 mutant allele under LD. We detected significant genetic interactions among CO1, CO2 and PPD1 genes on heading time, which were reflected in complex interactions at the transcriptional and protein levels. Loss-of-function mutations in PPD1 delayed heading more than combined co1 co2 mutations and, more importantly, PPD1 was able to perceive and respond to differences in photoperiod in the absence of functional CO1 and CO2 genes. Similarly, CO1 was able to accelerate heading time in response to LD in the absence of a functional PPD1. Taken together, these results indicate that PPD1 and CO1 are able to respond to photoperiod in the absence of each other, and that interactions between these two photoperiod pathways at the transcriptional and protein levels are important to fine-tune the flowering response in wheat. Author summaryAn understanding of the mechanisms involved in the regulation of wheat heading time is required to engineer more productive varieties better adapted to new or changing environments. A large proportion of wheats natural variation in heading time is determined by differences in genes controlling the photoperiodic response. In this study, we show that the wheat PHOTOPERIOD1 (PPD1) gene has a stronger effect on heading time than CONSTANS1 (CO1) and CO2 in the regulation of the photoperiodic response, and that complex genetic interactions among these genes are important to fine-tune heading time. Using loss-of-function mutants for both CO1 and CO2, we demonstrate that these genes are not required for PPD1 to perceive differences in photoperiod and regulate heading time. Similarly, we show that in the absence of PPD1, CO1 can accelerate heading time more than 60 days in response to long days. Our results indicate that each of these two wheat photoperiod pathways can respond to differences in photoperiod even in the absence of the other one. Differences in the relative importance of these two pathways and in their epistatic interactions have contributed to the diversity of photoperiodic responses observed in different grass species.

plant biology↗