Anticipation of respiratory outcomes in Acute Respiratory Distress Syndrome patients by clinical and metabolic signatures 

Area of Knowledge: LIFE SCIENCES

Group leaders

Santiago Marco, Institute for Bioengineering of Catalonia (IBEC) · 
Joan Ramón Masclans, Hospital del Mar ·


The clinical respiratory evolution of acute respiratory distress syndrome (ARDS) phenotypes is poorly understood. Identifying the metabolic pathways altered early with viral or bacterial infection and their association with failure of non-invasive ventilatory support strategies (NIS) is crucial to understanding ARDS pathophysiology, guiding clinical decisions, and determining important therapeutic targets for critical patients in the Intensive Care Unit. This study aimed to assess the specific alterations in metabolic pathways associated with NIS failure and the development of ARDS phenotype 2, related to poor outcomes. 

This project benefits from a large dataset collected in a previous project funded by LA MARARO de TV3 (Co-Respiromics), containing 396 ICU-admitted patients covering COVID epidemiological waves 1 to 5. The dataset contains 37 clinical variables and 140 metabolic markers for all the patients, collected at the 1, 3, and 5 from ICU admission.  

The objectives of the project are:  

1.- To confirm the predictive value of specific metabolic markers in determining NIS failure (need of mechanical ventilation/intubation) in a larger population of COVID-19 pneumonia patients at baseline (defined as 12 first hours of Intensive Care Unit admission). 

2.- To evaluate the predictive value of the specific biomarker level variations over time in determining NIS failure in the same population (increase or decrease at day 3 and day 5 of initiation of NIS). 

3.- To design a diagnostic and therapeutic making decision algorithm based on clinical, analytical and metabolomic information for daily clinical use. 

Job position description:

This interdisciplinary project aims to leverage advanced statistical methods and machine learning algorithms for the development of a therapeutic decision-making algorithm based on clinical and metabolomic data for Acute Respiratory Distress Syndrome patients in Intensive Care Units-  


  • Conduct high-quality research under the supervision of Dr. Santiago Marco (IBEC), and the clinical guidance of Dr. Joan R. Masclans (Hospital del Mar) 
  • Analyze large biomedical datasets using statistical methods. 
  • Design, implement and validate machine learning models for biomedical data. 
  • Discover metabolic markers and their association with clinical respiratory monitoring variables.  
  • Collaborate with clinicians, Hospital IT support, and other stakeholders. 
  • Publish findings in peer-reviewed journals and present at conferences. 


  • Master’s degree in bioinformatics, Biomedical Engineering, Data Science, or related field. 
  • Strong programming skills in Python, R, or MATLAB. 
  • Solid background in statistics and data analysis. 
  • Previous experience in machine learning and signal processing is highly desirable. 
  • Excellent written and oral communication skills in English. 

Desired Skills: 

  • Familiarity with health data acquisition and pre-processing. 
  • Vocation for research in life sciences and healthcare. 
  • Hands-on experience with biomedical datasets. 


  • Direct connection with clinical application case. 
  • Interdisciplinary research environment. 
  • Opportunities for career development and networking.