Resultados de investigación

The raw_data.zip file contains folders with the EEG data from all patients analyzed in .edf format. The clinical_data.csv file contains a dataset with the relevant clinical information for all patients in all recording sessions., Peer reviewed

Different methods for predicting lung transplant outcomes., Peer reviewed

Performance Comparison with Fair Machine Learning Baselines., Peer reviewed

Lung transplantation remains the only definitive treatment for end-stage respiratory failure; however, it has substantial post-operative mortality risk. Current methods like the Lung Transplant Risk Index offer limited predictive performance. This study introduces a novel topological feature engineering model to assess mortality risk. The objective is to improve predictive accuracy by capturing complex temporal patterns while ensuring interpretability. A retrospective cohort study was conducted using clinical data from lung transplant recipients. The model integrates static and time-dependent variables through topological feature extraction, enabling sequential risk updating at transplantation, ICU admission, and throughout early post-operative course. Performance was compared to established methods using a held-out test set. Metrics included accuracy, sensitivity, specificity, and AUC. Interpretability was assessed using Shapley Additive explanations. The proposed model demonstrated superior predictive performance compared to traditional clinical risk scores (LTRI, CCI) and standard machine learning models. On the test dataset, it achieved 87.4% accuracy, 84.1% sensitivity, and 89.6% specificity, with an absolute AUC gain of 0.08 over the best non-topological baseline (p < 0.001). The model consistently outperformed existing approaches across subgroups including age, underlying disease, and transplant type. Shapley analysis revealed that dynamic variables such as early post-operative oxygenation trends, immunosuppressive load, and inflammatory markers were among the most critical contributors to mortality risk. The integration of topological features significantly enhances prediction of post-transplant mortality risk. These findings highlight topological transformers as a valuable tool for precision medicine and clinical decision support., Peer reviewed

Effect Sizes with 95% Confidence Intervals for Top Predictive Features., Peer reviewed

Transcriptomic and scRNA-seq analyses identify metabolic switch and EMT induction upon pseudostarvation. Cells were treated for 48 hours with vehicle (Ctrl), MCM, and 20 μmol/L etomoxir (Eto). A, Gene expression profile as assessed by a carbohydrate metabolism PCR array in PDAC-354 cells. Heatmap showing only genes whose expression was significantly altered (n = 3). B, Left, PDAC-003 cells were treated as indicated and were then subjected to scRNA-seq (10X Genomics Chromium platform). Unsupervised clustering of viable PDAC cells exposed to Ctrl, MCM, or etomoxir, represented as Uniform Manifold Approximation and Projection (UMAP) plots. Different clusters are color-coded. Right, boxplots illustrating gene set enrichment results for the EMT and glycolysis (Hallmark data set) for different clusters in Ctrl vs. MCM and etomoxir treatment, respectively. Differences in enrichment scores between treatments were assessed using the Mann–Whitney U test. C, Expression of EMT hallmark signature and PPARD in single cancer cells (PDAC-002 and 021) displayed as unsupervised clusters and color-coded for allocated treatment. See also Supplementary Figs. S4 and S5., HORIZON EUROPE European Research Council (ERC) Seventh Framework Programme (FP7) National Science Foundation of China Fondazione AIRC per la ricerca sul cancro ETS (AIRC) Instituto de Salud Carlos III (ISCIII) Worldwide Cancer Research (WCR) Fundación Científica Asociación Española Contra el Cáncer (AECC) Fundación Fero (Fundació Fero) Cancer Council NSW (Cancer Council New South Wales) National Health and Medical Research Council (NHMRC) German Research Foundation Shanghai Municipal Education Commission (上海市教育委员会) Ministero della Salute (Italy Ministry of Health), Peer reviewed

Multikinase inhibitors have shown efficacy in endocrine neoplasms, and synergism with immune checkpoint inhibitors has been noted in other tumors. This is a prospective, multicenter, open-label, Simon two-stage optimal design, phase II study including patients with advanced and refractory endocrine and neuroendocrine neoplasms in six cohorts: lung well-differentiated neuroendocrine tumors, anaplastic thyroid cancer (ATC), adrenocortical carcinoma (ACC), pheochromocytoma/paraganglioma (PPGL), well-differentiated gastroenteropancreatic neuroendocrine tumors (GEP-NET), and grade 3 extrapulmonary neuroendocrine neoplasms. Patients received atezolizumab 1,200 mg intravenously every 3 weeks plus cabozantinib 40 mg/day orally until disease progression or unacceptable toxicity. The primary objective was the overall response rate (ORR) by RECIST 1.1. From October 2020 to December 2022, 93 patients were included. The ORR was 14.3% [95% confidence interval (CI), 1.8–42.8] in ATC (N = 14); 8.3% (95% CI, 1.0–27.0) in ACC (N = 24); 15.4% (95% CI, 1.9–45.5) in PPGL (N = 13), and 16.7% (95% CI, 4.7–37.4) in GEP-NET (N = 24). Lung well-differentiated neuroendocrine tumors and grade 3 extrapulmonary neuroendocrine neoplasms had no responses. The duration of response was 20.4 months in ATC, 13.1 months in ACC, 12.2 months in PPGL, and 15.8 months in GEP-NET. Survival rates at 12 months in ATC and ACC were 47.6% and 47.6%, respectively. No unexpected toxicity was observed. Cabozantinib and atezolizumab were safely administered and showed promising ORR, and preliminary long-term survival rates were observed in aggressive and pretreated ACC and ATC, which warrants further investigation., Roche (Roche Holding) Ipsen (Ipsen S.A.) GETNE, Peer reviewed

Panel (a) beta-lactamases, panel (b) ESBLs, panel (c) fluoroquinolones, panel (d) aminoglycosides., Peer reviewed

Panel (a) beta-lactamases conferring resistance to penicillins only (small and broad spectrum), panel (b) ESBL, panel (c) fluoroquinolones, and panel (d) aminoglycosides. The PSI is calculated with the following formula: where p corresponded to the relative frequency of gene type k (e.g. blaCTX-M-15) in reservoir n (e.g., ESBL-E. coli from humans), and q corresponded to the relative frequency of the same gene type in reservoir nx (e.g., ESBL-K. pneumoniae from humans). The numerator of the relative frequency was the count of each ARG type. The denominator of the relative frequency was the total number of genes of the corresponding ARG class. The PSI is a proportion, with 0 interpreted as no overlap, and 1 as perfect overlap in ARG type distributions between two reservoirs. 95% confidence intervals were calculated with 5,000 bootstrap iterations. PSI analysis for fosfomycin, trimethoprim, and sulfonamides ARGs are described in the supplement (S1 Fig 4 in S1 Appendix)., Peer reviewed

Nodes represent isolates and are grouped based on epidemiological setting, lines represent genetically similar isolate pairs. Panel (a) ESBL-producing E. coli (ESBL-Ec), panel (b) ESBL-producing K. pneumoniae (including seven K. variicola isolates) (ESBL-Kp), panel (c) ESBL-Ec Seville, panel (d) ESBL-Kp Seville. Grey boxes I and II are displayed enlarged in S1 Fig 5 in S1 Appendix, I: Panel A, II panel B., Peer reviewed