2021-03-04

040_ An integrative genomic and proteomic data approach to identify biomarkers related to cognition of bipolar disorder and schizophrenia patients

Research Question and Aims

The use of state-of-the-art proteomics and genomics techniques and the integration of both data modalities holds promise for the identification of reliable biomarkers in molecular psychiatry (Azmeraw T. Amare et al., 2017). In this research proposal that we will focus on neurocognition, usually effected in these patients, we have the following objectives:
1) Characterizing the genetic and longitudinal circulating proteomic signatures in bipolar patients (BD) that are potentially related to changes in neurocognitive performance over time.
2) Determining the correlation between genetic and circulating protein biomarkers with cross-sectional neurocognitive performance in BD and schizophrenia (SCZ).

Analytic Plan

Data
Phenotypic data:
State variables in PsyCourse related to several domains of cognition would be used in this study as target variables.

Biological data:
Proteomics: A first-pass screening (Mass Spectrometry) of ~450 proteins in the serum of ~120 BD patients and also ~100 proteins quantified by high-throughput antibody-based assay in serum of 113 SCZ and 125 BD patients (Papiol et al., 2019) are already available for analysis from the PsyCourse cohort.
Genomics: Raw genotypes (GSA chip) and imputed data from PsyCourse patients.

Association analyses & statistical approach
Using the aforementioned data, we will investigate:
1) The associations between polygenic load, protein (Mass Spectrometry) level changes and cognition variations over time in BD patients. These analyses will be adjusted for age, sex, treatment, population stratification (PCA or MDS components) and other relevant covariates in the context of repeated measures / linear mixed models.
2) The cross-sectional associations of polygenic load and a selected panel of ~100 serum proteins in 125 BD and 113 SCZ patients with cognitive performance will be assessed. These analyses will be adjusted for age, sex, treatment, population stratification (PCA or MDS components) and other relevant covariates in the context of linear/logistic models.

Resources needed

For those subjects with "Y" in v1_prot_id, v2_prot_id, v3_prot_id, v4_prot_id, v1_ab_prof_id, v2_ab_prof_id, v3_ab_prof_id:

Demographics
v1_sex

v1_age
v2_age
v3_age
v4_age

Neurocognition
v1_nrpsy_tmt_A_rt
v2_nrpsy_tmt_A_rt
v3_nrpsy_tmt_A_rt
v4_nrpsy_tmt_A_rt

v1_nrpsy_tmt_A_err
v2_nrpsy_tmt_A_err
v3_nrpsy_tmt_A_err
v4_nrpsy_tmt_A_err

v1_nrpsy_tmt_B_rt
v2_nrpsy_tmt_B_rt
v3_nrpsy_tmt_B_rt
v4_nrpsy_tmt_B_rt

v1_nrpsy_tmt_B_err
v2_nrpsy_tmt_B_err
v3_nrpsy_tmt_B_err
v4_nrpsy_tmt_B_err

v1_nrpsy_dgt_sp_frw
v2_nrpsy_dgt_sp_frw
v3_nrpsy_dgt_sp_frw
v4_nrpsy_dgt_sp_frw

v1_nrpsy_dgt_sp_bck
v2_nrpsy_dgt_sp_bck
v3_nrpsy_dgt_sp_bck
v4_nrpsy_dgt_sp_bck

v1_nrpsy_dg_sym
v2_nrpsy_dg_sym
v3_nrpsy_dg_sym
v4_nrpsy_dg_sym

v1_nrpsy_mwtb

Raw data on medication and drug use
200715_v4.1_psycourse_clin_raw_med_(visit1 to visit4).RData
200715_v4.1_psycourse_clin_raw_ill_drg_(visit1 to visit4).RData

Genomics: Raw genotypes (GSA chip) and imputed data from PsyCourse patients.

Proteomics: The first-pass screening (Mass Spectrometry) of ~450 proteins in the serum of ~120 BD patients and also the ~100 proteins quantified by high-throughput antibody-based assay in serum of 113 SCZ and 125 BD patients that are already available for analysis from the PsyCourse cohort.