All Institutions
Search all resources
Search for entire words only

Schizophrenia diagnosis and prognosis: finding the way to a personalized medicine

Faculty of Medicine, University of Coimbra
Project classification

Scientific area

5.1 Psychology


Psychology, special (including therapy for learning, speech, hearing, visual and other physical and mental disabilities)

Project description

Project title

Schizophrenia diagnosis and prognosis: finding the way to a personalized medicine

Scientific Coordinator's name:

António Macedo

Scientific Coordinator's e-mail:

Principal R&D Unit:

Department of Psychological Medicine, Faculty of Medicine, University of Coimbra

Other R&D Units involved in the project:

CNC.IBILI - Center for Neuroscience and Cell Biology (CNC) and the Institute for Biomedical Imaging and Life Sciences (IBILI)

Project keyword(s)

Schizophrenia Biomarker Diagnosis Prognosis Proteomics Metabolomics System modeling Predictive bioinformatics model

Short abstract and comments

Schizophrenia is a complex and chronic psychiatric disorder for which there is still no biomarker. The diagnosis of the disease is mainly based on clinical interview with no biomolecular support which can be used to increase diagnosis confidence or to guide prognosis. Moreover, medication resistant patients need to be subjected to long and unhelpful therapy trials before initiating clozapine, which is far from the goal of a more personalized and preventive medicine. Several potential genetic, protein, and metabolic biomarkers have been proposed in the literature or even commercialized. But in fact, their lack of robustness led to a marked withdrawal or not being clinically used due to poor diagnosis reliability. There is therefore a clear need for a comprehensive study which can combine all these approaches (clinical interview, proteins, metabolites and genes) into a single predictive model. This is precisely what this research project intends to do, gathering a clinical team with long experience in the diagnosis of thousands of patients, with a track record on psychiatric genetic and international consortiums besides being directly involved in the development of improved diagnostic procedures and interviews. The mass spectrometry team is a reference lab for the world leading MS seller, with a track record in proteomics, and several private research contracts on metabolite profiling and quantification. Finally the international collaboration with Pedro Beltrão from EMBL-EBI will ensure proper genetic and predictive model development. The research project will be based on extensive clinical characterization with state of the art diagnostic procedure with combined information from international referenced guidelines. The biobank team has experience as certified lab for international neurogenerative disorders, as the Joint Programming for Neurodegenerative Diseases (JPND), devoted to the standardization of biomarkers (BIOMARKAPD and DEMTEST). Each individual sample will be also characterized with common blood tests and complemented with immunophenotyping using a panel of markers and flow cytometry in collaboration with the Portuguese Blood Institute. Then, three parallel studies will be performed: i) proteomics, ii) metabolomics, and iii) targeted gene analysis. The proteomics screening will analyze not only serum and plasma (equalized with ProteoMiner) but will also enrich the samples for phosphopeptides and cysteine modificated proteins (using iodoTMT). Furthermore, a detailed and comprehensive analysis will be performed on Peripheral Blood Mononuclear Cells, as there is an increased knowledge on the association of SCZ with the immune system. The metabolomics screening will cover the three previously mentioned samples, with the novelty of using a SWATH-MS approach to produce screening information besides quantitative data. Moreover, a targeted analysis on neurotransmitter metabolism will be performed in order to elucidate either earlier neurotransmitter imbalance or therapy modulation. The genetic approach will be targeted on previously identified genes related with SCZ although they have not been able to perform the diagnosis alone, our rationale is that it will be the combination of different parameters with different weighting factor that can create the predictive model. This is the major goal of the bioinformatics team. These multiple factors will be validated with different populations from the country’s largest hospital. The novelty of the project is not on each individual task, but the depth of the screening performed in each patient, and the overall capacity to integrate the information from four completely independent screens. It is expected to deliver a predictive model which based on a reduced panel of biomarkers will provide enough information to be used as diagnosis or prognosis for SCZ, alone or in combination with the clinical interview. Moreover, already diagnosed patients can benefit by predicting if their molecular signature indicates them as responder or non-responders of the current therapy prescribed, along with potential disease progression. We believe that besides the enormous clinical outcome of this ambitious but feasible project, there is also a huge economical impact, as SCZ patients frequently have work impairment and difficult integration in the society.

Potential uses/indications

To our knowledge this is the first time that a comprehensive analysis combining these four major data sources is performed, along with the expertise of predictive modeling. This model will be challenged with a new set of samples not previously used to construct the model (validation on task 6). These samples are important to perform adjustments, including the analysis of first relatives to exclude confounding factors. Finally, the model will be tested on double-blind samples, including first episode patients, first relative patients, BD, well characterized drug responders and non-responders. This project will generate a massive knowledge on a well characterized set of patients, filling the gap from previous non-related studies, and will increase our current molecular knowledge on SCZ pathophysiology. Most importantly, it is expected to deliver a molecular and interview based predictive model which can be used both for diagnosis and prognosis.



Partner Status: Seeking Partners?


Grant number (QREN, FP7, Eureka, etc)


Last edited on

2015-09-02 11:24:51

Health Cluster Portugal All rights reserved
© 2024 Health Cluster Portugal All rights reserved