Biology subjects

Morilla, I.

Publications and source records attributed to Morilla, I..

8 recordsLinked to original sources

bioRxiv2026

What does domain-specific pretraining buy? A 170,000-parameter plant peptide language model against the ESM-2 scaling curve

Small signalling peptides (SSPs) coordinate development and stress response in plants, and the computational tools that identify them represent sequences using general-purpose protein language models trained across all known proteins. How much a representation learned on the SSP sequence space itself is worth, measured against generalist embeddings of varying size, has not to our knowledge been quantified. We pretrain PhytoPepBERT, a 169,567-parameter masked language model implemented from scratch with no deep-learning framework, on 1,469 plant SSP precursors retrieved from UniProtKB spanning five peptide families (RALF, PSK, CLE, PEP, RGF) and a length-matched negative class of non-SSP plant proteins. Sequences are clustered at 70% alignment identity and every cross-validation fold is grouped by cluster, so no homologue of a held-out sequence appears in training. Domain pretraining is worth a great deal relative to where it starts. Family classification by k-nearest neighbours over frozen embeddings reaches 85.2% +/- 2.4% against 58.0% +/- 4.3% for an architecturally identical, identically initialised untrained backbone (chance 16.7%), and against 75.1% for dipeptide composition and 67.6% for BLOSUM62. The advantage is largest where the backbone stays frozen: linear probing gives 82.9% against 67.6% untrained, while full fine-tuning narrows the gap to 85.4% against 83.6%. Masked-residue accuracy is 20.5% against 3.4% untrained, elevated across all five peptide families and lowest on the non-SSP negatives. Benchmarked against ESM-2 at four scales under an identical protocol, those 169,567 domain-pretrained parameters buy approximately what 7.5M generalist parameters buy: PhytoPepBERT ties esm2_t6_8M_UR50D (86.1% +/- 2.8\%, two of five paired folds) and is overtaken thereafter, by 2.5 points at 35M, 7.0 at 150M and 11.6 at 650M, losing every paired fold above the smallest checkpoint. TAPE, the second representation the reference pipeline uses, leads by 3.6 points at 92M. Generalist scale therefore continues to pay on this task well beyond the point where a domain model of this size can follow. We report an exchange rate rather than a victory: domain pretraining delivers roughly a forty-fold parameter efficiency against the generalist curve, and nothing beyond it.

plant biology↗

bioRxiv2026

Foundational characterization of tomato fruit RALF peptides reveals structural and functional specialization within the SlRALF

Rapid Alkalinization Factor (RALF) peptides regulate plant growth and cell wall signaling, but their roles in fruit development remain unclear. Here, we characterized tomato (Solanum lycopersicum) fruit-associated RALF peptides and their interactions with leucine-rich repeat extensins (LRXs). Expression analyses identified SlRALF5, SlRALF7, and SlRALF10 as the main fruit-expressed RALFs. SlRALF10 was associated with early fruit development, whereas SlRALF5 and SlRALF7 remained expressed during ripening. Sequence analyses showed that SlRALF5/7 retain conserved motifs of canonical RALFs, while SlRALF10 displays divergent structural features and altered charge distribution. Synthetic SlRALF5 and SlRALF7 inhibited root growth and induced extracellular alkalinization, whereas SlRALF10 lacked both activities. Co-immunoprecipitation assays showed that all three peptides interact with the fruit-expressed proteins SlLRX2 and SlLRX5. Structural modeling predicted distinct electrostatic properties for the SlLRX5/SlRALF10 complex compared with SlRALF5. These results reveal structural and functional specialization among tomato fruit RALF peptides and suggest that distinct SlRALFs may differentially respond to cell wall remodeling during fruit development and ripening.

plant biology↗

bioRxiv2025

Multimodal Learning Reveals Plants' Hidden Sensory Integration Logic

Plants integrate complex environmental signals through interconnected molecular networks, but the fundamental rules governing this sensory integration remain unknown. Studying tomato roots interacting with fungal symbionts, we discover how microbial effectors systematically reprogram plant sensory systems by coordinating transcriptional, metabolic, and phenotypic responses. Our multimodal analysis not only confirmed prior experimental findings through purely computational means, but also revealed novel integration hubs where sensory pathways converge. This dual validation approach establishes iron homeostasis rewiring through citrate-mediated redox control. Next, targeted suppression of jasmonate defences. Thus, nuclear splicing isolation from metabolic noise. These findings establish a new paradigm for understanding plant-microbe communication, showing how symbionts exploit latent hubs where sensory pathways converge. The discovered integration logic provides both fundamental insights into plant perception and concrete targets for engineering stress-resilient crops.

plant biology↗

bioRxiv2025

Arabidopsis lines with modified ascorbate concentrations reveal a link between ascorbate and auxin biosynthesis

Ascorbate is the most abundant water-soluble antioxidant in plants, and it is an essential molecule for normal plant development. It is present in all green plants, with very different concentrations in different plant species. While ascorbate accumulation is a trait of nutritional, and therefore, agronomical interest, the impact of different concentrations over cellular homeostasis remains elusive. In order to shed light over this question, we leveraged Arabidopsis lines with very low ascorbate (vtc2 mutant with 20% of WT ascorbate levels), and low ascorbate concentration (vtc4 mutant with 65% of WT levels), and we generated a line that accumulates 165% of WT levels (vtc2/OE-VTC2). An 80% reduction of ascorbate increased the expression of genes implicated in defense against pathogens, but repressed genes associated with abiotic stress responses. Unexpectedly, lines with increased (165% of WT) and decreased (65% of WT) ascorbate levels shared 85% of induced transcription factors and the GO terms associated with their transcriptional programs. Among the group of genes whose expression is positively correlated with ascorbate content, we identified TAA1/WEI8, a gene encoding a tryptophan aminotransferase that catalyzes the first step of auxin biosynthesis. Using a combination of genetic and pharmacological approaches in fluorescent and histochemical reporter lines for auxin biosynthesis and signaling activity, we revealed that TAA1- and TAA1-RELATED2 (TAR2)-mediated auxin biosynthesis is necessary for plants to cope with increased ascorbate concentration in a light-dependent manner, revealing a new layer of complexity in the regulatory landscape of redox homeostasis.

plant biology↗

bioRxiv2024

S2-PepAnalyst: A Web Tool for Predicting Plant Small Signalling Peptides

Small signalling peptides mediate cell-to-cell communication playing essential roles in plant growth, development, and stress responses. They specifically bind to the extracellular domain of receptors to trigger biochemical and physiological responses. Despite their significance, accurately identifying novel signalling peptides remains challenging due to their structural diversity, low abundance, and highly specific expression patterns. Here, we present S{superscript 2}-PepAnalyst, a web tool integrating plant-specific datasets and machine learning to predict SSPs with 99.5% accuracy and low false-negative rates. S{superscript 2}-PepAnalyst outperforms existing tools (e.g., SignalP 6.0) by combining protein language models, geometric-topological analysis, and reinforcement learning, enabling robust classification of small signalling peptide families (e.g., CLE, RALF). The tool is freely available at https://www.s2-pepanalyst.uma.es.

plant biology↗

bioRxiv2022

Proteome dynamics of COVID-19 severity learnt by a graph convolutional network of multi-scale topology

Many efforts have been recently done to characterise the molecular mechanisms of COVID-19 disease. These efforts resulted in a full structural identification of ACE2 as principal receptor of the Sars-CoV-2 spike protein in the cell. However, there are still important open questions related to other proteins involved in the progression of the disease. To this end, we have modelled the plasma proteome of 384 COVID patients. The model calibrated proteins measures at three time tags and make also use of the detailed clinical evaluation outcome of each patient after their hospital stay at day 28. Our analysis is able to discriminate severity of the disease by means of a metric based on available WHO scores of disease progression. Then, we identify by topological vectorisation those proteins shifting the most in their expression depending on that severity classification. Finally, the extracted topological invariants respect the protein expression at different times were used as base of a graph convolutional network. This model enabled the dynamical learning of the molecular interactions produced between the identified proteins.

systems biology↗

bioRxiv2019

Singular manifolds of proteomic drivers to model the evolution of inflammatory bowel disease status

The conditions who denotes the presence of an immune disease are often represented by interaction graphs. These informative, but complex structures are susceptible to being perturbed at different levels. The mode in which that perturbation occurs is still of utmost importance in areas such as reprogramming therapeutics. In this sense, the overall graph architecture is well characterise by module identification. Topological overlap-related measures make possible the localisation of highly specific module regulators that can perturb other nodes, potentially causing the entire system to change behaviour or collapse. We provide a geometric framework explaining such situations in the context of inflammatory bowel diseases (IBD). IBD are important chronic disorders of the gastrointestinal tract which incidence is dramatically increasing worldwide. Our approach models different IBD status as Riemannian manifolds defined by the graph Laplacian of two high throughput proteome screenings. Identifies module regulators as singularities within the manifolds (the so-called singular manifolds). And reinterprets the characteristic IBD nonlinear dynamics as compensatory responses to perturbations on those singularities. Thus, we could control the evolution of the disease status by reconfiguring particular setups of immune system to an innocuous target state.

systems biology↗

bioRxiv2018

Functional systemic CD4 immunity is a differential factor for clinical responses to PD-L1/PD-1 blockade therapy in lung cancer

The majority of lung cancer patients progressing from conventional therapies are refractory to PD-L1/PD-1 blockade monotherapy. Here we show that baseline systemic CD4 immunity is a differential factor for clinical responses. Patients with functional systemic CD4 T cells included all objective responders and could be identified before the start of therapy by having a high proportion of memory CD4 T cells. In these patients CD4 T cells possessed significant proliferative capacities, co-expressed low levels of PD-1/LAG-3 and were responsive to PD-1 blockade ex vivo and in vivo. In contrast, patients with dysfunctional systemic CD4 immunity did not respond even though they had circulating lung cancer antigen-specific CD4 and CD8 T cells. CD4 T cells in these patients proliferated very poorly, strongly co-upregulated PD-1/LAG-3 after stimulation, and were largely refractory to PD-1 monoblockade ex vivo and in vivo. Although baseline CD8 immunity was strongly dysfunctional in all patients before the start of therapy, it only recovered in patients with functional CD4 immunity. CD4 and CD8 T cell dysfunctionality was caused by PD-1/LAG-3 co-expression, which could be reverted by PD-1/LAG-3 co-blockade. Patients with functional CD4 immunity and PD-L1 tumor positivity exhibited response rates of 70%, highlighting the contribution of CD4 immunity for efficacious PD-L1/PD-1 blockade therapy.

cell biology↗