Pairing interacting protein sequences using masked language modeling.

Predicting which proteins interact together from amino acid sequences is an important task. We develop a method to pair interacting protein sequences which leverages the power of protein language models trained on multiple sequence alignments (MSAs), such as MSA Transformer and the EvoFormer module of AlphaFold. We formulate the problem of pairing interacting partners among the paralogs of two protein families in a differentiable way. We introduce a method called Differentiable Pairing using Alignment-based Language Models (DiffPALM) that solves it by exploiting the ability of MSA Transformer to fill in masked amino acids in multiple sequence alignments using the surrounding context. MSA Transformer encodes coevolution between functionally or structurally coupled amino acids within protein chains. It also captures inter-chain coevolution, despite being trained on single-chain data. Relying on MSA Transformer without fine-tuning, DiffPALM outperforms existing coevolution-based pairing methods on difficult benchmarks of shallow multiple sequence alignments extracted from ubiquitous prokaryotic protein datasets. It also outperforms an alternative method based on a state-of-the-art protein language model trained on single sequences. Paired alignments of interacting protein sequences are a crucial ingredient of supervised deep learning methods to predict the three-dimensional structure of protein complexes. Starting from sequences paired by DiffPALM substantially improves the structure prediction of some eukaryotic protein complexes by AlphaFold-Multimer. It also achieves competitive performance with using orthology-based pairing.

Generative power of a protein language model trained on multiple sequence alignments.

Computational models starting from large ensembles of evolutionarily related protein sequences capture a representation of protein families and learn constraints associated to protein structure and function. They thus open the possibility for generating novel sequences belonging to protein families. Protein language models trained on multiple sequence alignments, such as MSA Transformer, are highly attractive candidates to this end. We propose and test an iterative method that directly employs the masked language modeling objective to generate sequences using MSA Transformer. We demonstrate that the resulting sequences score as well as natural sequences, for homology, coevolution, and structure-based measures. For large protein families, our synthetic sequences have similar or better properties compared to sequences generated by Potts models, including experimentally validated ones. Moreover, for small protein families, our generation method based on MSA Transformer outperforms Potts models. Our method also more accurately reproduces the higher-order statistics and the distribution of sequences in sequence space of natural data than Potts models. MSA Transformer is thus a strong candidate for protein sequence generation and protein design.

Impact of phylogeny on structural contact inference from protein sequence data.

Local and global inference methods have been developed to infer structural contacts from multiple sequence alignments of homologous proteins. They rely on correlations in amino acid usage at contacting sites. Because homologous proteins share a common ancestry, their sequences also feature phylogenetic correlations, which can impair contact inference. We investigate this effect by generating controlled synthetic data from a minimal model where the importance of contacts and of phylogeny can be tuned. We demonstrate that global inference methods, specifically Potts models, are more resilient to phylogenetic correlations than local methods, based on covariance or mutual information. This holds whether or not phylogenetic corrections are used, and may explain the success of global methods. We analyse the roles of selection strength and of phylogenetic relatedness. We show that sites that mutate early in the phylogeny yield false positive contacts. We consider natural data and realistic synthetic data, and our findings generalize to these cases. Our results highlight the impact of phylogeny on contact prediction from protein sequences and illustrate the interplay between the rich structure of biological data and inference.

Clinical complexity and diabetes: a multidimensional approach for the management of cardiorenal metabolic syndrome.

Diabetes mellitus (DM) is one of the fastest-growing health emergencies of the 21st century, and one of the chronic diseases with the highest socio-economic impact on health care systems. DM is the main cause of chronic kidney disease, and is associated with a significant increase in cardiovascular risk and clinical and care complexity. The presence of a constellation of cardiac, metabolic, and renal diseases, in a complex patient with DM, constitutes the CardioRenal Metabolic Syndrome (CRMS). The management of these patients should include a paradigm shift from a reactive strategy to a proactive approach, and the integration of territorial, hospital and social assistance services according to the Chronic Care Model (CCM). Complexity science suggests an alternative model in which disease and health arise from complex, dynamic, and unique interactions among the different components of the overall system. The hospital should be viewed as a highly specialized hub of the chronic care system, which interacts with the outpatient specialist and primary care. In order to create effective communication among territorial care units and highly specialized hospitals, levels of clinical complexity are here proposed and included in a multidimensional management model for the complex patient with diabetes and cardiorenal comorbidity.

Protein language models trained on multiple sequence alignments learn phylogenetic relationships.

Self-supervised neural language models with attention have recently been applied to biological sequence data, advancing structure, function and mutational effect prediction. Some protein language models, including MSA Transformer and AlphaFold's EvoFormer, take multiple sequence alignments (MSAs) of evolutionarily related proteins as inputs. Simple combinations of MSA Transformer's row attentions have led to state-of-the-art unsupervised structural contact prediction. We demonstrate that similarly simple, and universal, combinations of MSA Transformer's column attentions strongly correlate with Hamming distances between sequences in MSAs. Therefore, MSA-based language models encode detailed phylogenetic relationships. We further show that these models can separate coevolutionary signals encoding functional and structural constraints from phylogenetic correlations reflecting historical contingency. To assess this, we generate synthetic MSAs, either without or with phylogeny, from Potts models trained on natural MSAs. We find that unsupervised contact prediction is substantially more resilient to phylogenetic noise when using MSA Transformer versus inferred Potts models.

Effectiveness and safety of concurrent beta-blockers and inhaled bronchodilators in COPD with cardiovascular comorbidities.

Chronic obstructive pulmonary disease (COPD) is the most common chronic respiratory disease and its prevalence is increasing worldwide, in both industrialised and developing countries. Its prevalence is ∼5% in the general population and it is the fourth leading cause of death worldwide. COPD is strongly associated with cardiovascular diseases; in fact, ∼64% of people suffering from COPD are treated for a concomitant cardiovascular disease and approximately one in three COPD patients die as a consequence of cardiovascular diseases.Inhaled bronchodilators might have adverse cardiovascular effects, including ischaemic events and arrhythmias, and beta-blockers might adversely influence the respiratory symptoms and the response to bronchodilators. For these reasons, it is important to know the safety profiles and the possible interactions between these two classes of drug, in order to prescribe them with greater awareness.In this article, we review the literature about the epidemiology of COPD, its association with cardiovascular diseases, and the safety of concurrent use of inhaled bronchodilators and beta-blockers, as a tool for improving the approach to complex therapies in clinical practice.

Hypertonic Saline in Conjunction with High-Dose Furosemide Improves Dose-Response Curves in Worsening Refractory Congestive Heart Failure.

INTRODUCTION: Diuretic responsiveness in patients with chronic heart failure (CHF) is better assessed by urine production per unit diuretic dose than by the absolute urine output or diuretic dose. Diuretic resistance arises over time when the plateau rate of sodium and water excretion is reached prior to optimal fluid elimination and may be overcome when hypertonic saline solution (HSS) is added to high doses of furosemide. METHODS: Forty-two consecutively hospitalized patients with refractory CHF were randomized in a 1:1:1 ratio to furosemide doses (125 mg, 250 mg, 500 mg) so that all patients received intravenous furosemide diluted in 150 ml of normal saline (0.9%) in the first step (0-24 h) and the same furosemide dose diluted in 150 ml of HSS (1.4%) in the next step (24-48 h) as to obtain 3 groups as follows: Fourteen patients receiving 125 mg (group 1), fourteen patients receiving 250 mg (group 2), and fourteen patients receiving 500 mg (group 3) of furosemide. Urine samples of all patients were collected at 30, 60, and 90 min, and 3, 4, 5, 6, 8, and 24 h after infusion. Diuresis, sodium excretion, osmolality, and furosemide concentration were evaluated for each urine sample. RESULTS: After randomization, 40 patients completed the study. Two patients, one in group 2 and one in group 3 dropped out. Patients in group 1 (125 mg furosemide) had a mean age of 77 ± 17 years, 43% were male, 6 (43%) had heart failure with a preserved ejection fraction (HFpEF), and 64% were in New York Heart Association (NYHA) class IV; the mean age of patients in group 2 (250 mg furosemide) was 80 ± 8.1 years, 15% were male, 5 (38%) had HFpEF, and 84% were in NYHA class IV; and the mean age of patients in group 3 (500 mg furosemide) was 73 ± 12 years, 54% were male, 6 (46%) had HFpEF, and 69% were in NYHA class IV. HSS added to furosemide increased total urine output, sodium excretion, urinary osmolality, and furosemide urine delivery in all patients and at all time points. The percentage increase was 18,14, and 14% for urine output; 29, 24, and 16% for total sodium excretion; 45, 34, and 20% for urinary osmolarity; and 27, 36, and 32% for total furosemide excretion in groups 1, 2, and 3, respectively. These findings were translated in an improvement in the furosemide dose-response curves in these patients. CONCLUSION: These results may serve as new pathophysiological basis for HSS use in the treatment of refractory CHF.

Troponin I release after intravenous treatment with high furosemide doses plus hypertonic saline solution in decompensated heart failure trial (Tra-HSS-Fur).

BACKGROUND: High values of cardiac troponin in acute decompensated congestive heart failure (ADHF) identify patients at higher risk and worsened prognosis. A cardiac troponin increase during therapy indicates the need for more appropriate intervention, aimed at compensating cardiac disease and effectively minimizing myocardial wall stress and subsequent cytolysis. This study evaluated the effects of an intravenous high dose of furosemide with (group A) or without small volume hypertonic saline solution (HSS) (group B) on myocardial cytolysis in patients with ADHF. METHODS: A total of 248 consecutive patients with ADHF (148 men, mean age 74.9 ± 10.9 years) were randomly assigned to group A or B. Plasma levels of cardiac troponin-I, brain natriuretic peptide, glomerular filtration rate by Modification of Diet in Renal Disease formula, bioelectrical impedance analysis measurements, and delta pressure/delta time (dP/dt) rate were observed on admission and discharge for all patients. RESULTS: We observed a significant reduction of cardiac troponin in both groups and a significant improvement in renal function, hydration state, pulmonary capillary wedge pressure (P < .0001), end diastolic volume (P < .01), ejection fraction (P < .01), and dP/dt (P < .004) in group A. We also observed a significant reduction in body weight (64.4 vs 75.8 kg) (P < .001), cardiac troponin I (0.02 vs 0.31 ng/mL) (P < .0001) and brain natriuretic peptide (542 vs 1,284 pg/mL) (P < .0001), and hospitalization time (6.25 vs 10.2 days) (P < .0001) in the HSS group. CONCLUSIONS: These data demonstrate that intravenous high doses of furosemide do not increase myocardial injury and, in addition, when associated to HSS, significantly reduce cardiac troponin I release. This behavior is mirrored by the achievement of improved hemodynamic compensation at echocardiography and body hydration normalization.

Large hiatal hernia at chest radiography in a woman with cardiorespiratory symptoms.

Hiatal hernia (HH) is a frequent entity. Rarely, it may exert a wide spectrum of clinical presentations mimicking acute cardiovascular events such as angina-like chest pain until manifestations of cardiac compression that can include postprandial syncope, exercise intolerance, respiratory function, recurrent acute heart failure, and hemodynamic collapse. A 69-year-old woman presented to the emergency department complaining of fatigue on exertion, cough, and episodes of restrosternal pain with less than 1 hour of duration. Her medical history only included some episodes of bronchitis and no history of hypertension. The 12-lead electrocardiogram demonstrated sinus rhythm with right bundle-branch block. Laboratory tests, including cardiac troponin I, were within normal reference values. Chest radiography showed no significant pulmonary alterations and revealed in mediastinum a huge abnormal shadow overlapping the right heart compatible with a gastric bubble.The gastroscopy confirmed a large HH. A 2-dimensional transthoracic echocardiogram, using all standard and modified apical and parasternal views, revealed an echolucent mass, compatible with HH, compressing the right atrium. Also, it showed an altered left ventricular relaxation and a mild increase of pulmonary artery pressure (35 mm Hg). Spirometry showed a mild obstruction of the small airways, whereas coronary angiography showed normal coronary arteries. We concluded that the patient's symptomatology was related to the compressive effects of the large hiatal ernia, a neglected cause of cardiorespiratory symptoms. The surgical repair of HH was indicated.