Among crop fruit trees, the apricot (Prunus armeniaca) provides an excellent model to study divergence and adaptation processes. Here, we obtain nearly 600 Armeniaca apricot genomes and four high-quality assemblies anchored on genetic maps. Chinese and European apricots form two differentiated gene pools with high genetic diversity, resulting from independent domestication events from distinct wild Central Asian populations, and with subsequent gene flow. A relatively low proportion of the genome is affected by selection. Different genomic regions show footprints of selection in European and Chinese cultivated apricots, despite convergent phenotypic traits, with predicted functions in both groups involved in the perennial life cycle, fruit quality and disease resistance. Selection footprints appear more abundant in European apricots, with a hotspot on chromosome 4, while admixture is more pervasive in Chinese cultivated apricots. Our study provides clues to the biology of selected traits and targets for fruit tree research and breeding.

Patient recruitment remains a major bottleneck in clinical trials, calling for scalable and automated solutions. We present TrialMatchAI, an AI-powered recommendation system that automates patient-to-trial matching by processing heterogeneous clinical data, including structured records and unstructured physician notes. Built on fine-tuned, open-source large language models (LLMs) within a retrieval-augmented generation framework, TrialMatchAI ensures transparency and reproducibility and maintains a lightweight deployment footprint suitable for clinical environments. The system normalizes biomedical entities, retrieves relevant trials using a hybrid search strategy combining lexical and semantic similarity, re-ranks results, and performs criterion-level eligibility assessments using medical Chain-of-Thought reasoning. This pipeline delivers explainable outputs with traceable decision rationales. In real-world validation, 92% of oncology patients had at least one relevant trial retrieved within the top 20 recommendations. Evaluation across synthetic and real clinical datasets confirmed state-of-the-art performance, with expert assessment validating over 90% accuracy in criterion-level eligibility classification - particularly excelling in biomarker-driven matches. Designed for modularity and privacy, TrialMatchAI supports Phenopackets-standardized data, enables secure local deployment, and allows seamless replacement of LLM components as more advanced models emerge. By enhancing efficiency, interpretability and offering lightweight, open-source deployment, TrialMatchAI provides a scalable solution for AI-driven clinical trial matching in precision medicine.

Calcium ions (Ca²⁺) play important and diverse roles in the regulation of autophagy, cell death and differentiation. Here, we investigated the impact of Ca²⁺ in regulating acute promyelocytic leukemia (APL) cell fate in response to the anti-cancer agent all-trans retinoic acid (ATRA). We observed that ATRA promotes calcium entry through store-operated calcium (SOC) channels into acute promyelocytic leukemia (APL) cells. This response is associated with changes in the expression profiles of ORAI1 and STIM1, two proteins involved in SOC channels activation, as well as with a significant upregulation of several key proteins associated to calcium signaling. Moreover, ATRA treatment of APL cells led to a significant activation of calcium/calmodulin-dependent protein kinase kinase 2 (CAMKK2) and its downstream effector AMP-activated protein kinase (AMPK), linking Ca²⁺ signaling to autophagy. Pharmacological inhibition of SOC channels and CAMKK2 enhanced ATRA-induced cell differentiation and death. Altogether, our results unravel an ATRA-elicited signaling pathway that involves SOC channels/CAMKK2 activation, induction of autophagy, inhibition of cellular differentiation and suppression of cell death. We suggest that SOC channels and CAMKK2 may constitute novel drug targets for potentiating the anti-cancer effect of ATRA in APL patients.

We have estimated the prevalence of the different viral subtypes between January 2012 and December 2016 in HIV-1-infected patients of the Aquitaine region (south-west part of France) who had a routine HIV-1 genotype resistance testing (GRT) centralized at the Bordeaux University Hospital. GRT was performed on viral RNA (1784 samples) before treatment initiation or at failure whereas proviral DNA was used as template (1420 samples) in the event of a treatment switch in patients with viral load below 50 copies/mL. Pol and integrase sequences were obtained; subtypes, circulating recombinant forms (CRF) and unique recombinant forms (URF) were assigned by combining the results of SCUEAL, REGA, COMET and HIV BLAST. Globally, subtype B was predominant with 71.7% while non-B subtypes accounted for 28.3%. Within the non-B viruses, CRF02_AG was the most prominent (11.6%) followed by non-B non-URF (13.5%), A, CRF01_AE, G, CRF06_cpx, F, C, D, H, J and finally URF (3.2%). The analysis of the two compartments separately showed that RNA exhibits higher percentages of non-B viruses than DNA. This study reveals a high degree of diversity of HIV-1 non-B subtype strains in Aquitaine, with an increasing prevalence of CRF02_AG and URF in the population investigated for viral RNA id est including more recently detected HIV-1 infected patients. Future studies should attempt to identify the transmission clusters while paying special attention to URF, since they seem to be increasing in the population and could potentially host CRF.

Targeted therapies have greatly improved cancer patient prognosis. For instance, chronic myeloid leukemia is now well treated with imatinib, a tyrosine kinase inhibitor. Around 80% of the patients reach complete remission. However, despite its great efficiency, some patients are resistant to the drug. This heterogeneity in the response might be associated with pharmacokinetic parameters, varying between individuals because of genetic variants. To assess this issue, next-generation sequencing of large panels of genes can be performed from patient samples. However, the common problem in pharmacogenetic studies is the availability of samples, often limited. In the end, large sequencing data are obtained from small sample sizes; therefore, classical statistical analyses cannot be applied to identify interesting targets. To overcome this concern, here, we described original and underused statistical methods to analyze large sequencing data from a restricted number of samples. To evaluate the relevance of our method, 48 genes involved in pharmacokinetics were sequenced by next-generation sequencing from 24 chronic myeloid leukemia patients, either sensitive or resistant to imatinib treatment. Using a graphical representation, from 708 identified polymorphisms, a reduced list of 115 candidates was obtained. Then, by analyzing each gene and the distribution of variant alleles, several candidates were highlighted such as UGT1A9, PTPN22, and ERCC5. These genes were already associated with the transport, the metabolism, and even the sensitivity to imatinib in previous studies. These relevant tests are great alternatives to inferential statistics not applicable to next-generation sequencing experiments performed on small sample sizes. These approaches permit to reduce the number of targets and find good candidates for further treatment sensitivity studies.

This study aimed at searching for the enzymes that are responsible for the higher hydroxylation of flavonols serving as UV-honey guides for pollinating insects on the petals of Asteraceae flowers. To achieve this aim, an affinity-based chemical proteomic approach was developed by relying on the use of quercetin-bearing biotinylated probes, which were thus designed and synthesized to selectively and covalently capture relevant flavonoid enzymes. Proteomic and bioinformatic analyses of proteins captured from petal microsomes of two Asteraceae species (Rudbeckia hirta and Tagetes erecta) revealed the presence of two flavonol 6-hydroxylases and several additional not fully characterized proteins as candidates for the identification of novel flavonol 8-hydroxylases, as well as relevant flavonol methyl- and glycosyltransferases. Generally speaking, this substrate-based proteome profiling methodology constitutes a powerful tool for the search for unknown (flavonoid) enzymes in plant protein extracts.