Search

Background At present, no predictive markers for Major Depressive Disorder (MDD) exist. The search for such markers has been challenging due to clinical and molecular heterogeneity of MDD, the lack of statistical power in studies and suboptimal statistical tools applied to multidimensional data. Machine learning is a powerful approach to mitigate some of these limitations. Methods We aimed to identify the predictive markers of recurrent MDD in the elderly using peripheral whole blood from the Sydney Memory and Aging Study (SMAS) (N = 521, aged over 65) and adopting machine learning methodology on transcriptome data. Fuzzy Forests is a Random Forests-based classification algorithm that takes advantage of the co-expression network structure between genes; it allows to alleviate the problem of p >> n via reducing the dimensionality of transcriptomic feature space. Results By adopting Fuzzy Forests on transcriptome data, we found that the downregulated TFRC (transferrin receptor) can predict recurrent MDD with an accuracy of 63%. Limitations Although we corrected our data for several important confounders, we were not able to account for the comorbidities and medication taken, which may be numerous in the elderly and might have affected the levels of gene transcription. Conclusions We found that downregulated TFRC is predictive of recurrent MDD, which is consistent with the previous literature, indicating the role of the innate immune system in depression. This study is the first to successfully apply Fuzzy Forests methodology on psychiatric condition, opening, therefore, a methodological avenue that can lead to clinically useful predictive markers of complex traits.

Background DISC1 imaging genetics studies in healthy controls, schizophrenia, and bipolar disorder cases have revealed morphological changes in brain regions involved in the pathophysiology of psychiatric disease including the hippocampus, anterior cingulate cortex (ACC), and the striatum. However, many of these studies have yielded discordant findings so there is a need for replication. Furthermore, despite evidence from human genetic studies and animal models implicating DISC1 in major depressive disorder (MDD), a DISC1 imaging genetics study in MDD cases has yet to be published. Thus, using neuroimaging data from MDD cases and a large sample of healthy controls we aimed to identify morphological changes representing neurobiological mechanisms underlying the association between DISC1 and MDD. Methods We utilized structural magnetic resonance imaging (sMRI) data from 512 healthy controls and 171 current MDD (SCID interview) cases, each with genotype data for non-synonymous DISC1 SNPs rs3738401, rs6675281, and rs821616. Results Region of interest analyses failed to reveal DISC1 -associated morphological changes in the hippocampus, ACC, or striatum in MDD patients and healthy controls. Whole brain exploratory analyses identified a nominally significant cluster mapping to the border of the precentral and postcentral gyri associated with rs821616 in healthy controls only ( p (uncorrected) Limitations We focused our analyses exclusively on three, but previously heavily studied, SNPs in DISC1 . Conclusions Our findings suggest that morphological changes in the hippocampus, ACC, and/or striatum of MDD patients do not represent neurobiological mechanisms underlying the association between DISC1 and MDD. However, we urge replication in independent samples of MDD cases.

Residual impairments in cognitive domains including attention, memory, executive function, and social perception occur in about 40% of patients with Major Depressive Disorder (MDD) following remission from acute episodes (Conradi et al., 2011), and are associated with reductions in pre-morbid functioning and higher unemployment rates (Baune et al., 2010; Evans et al., 2014). Cognitive impairment in MDD may represent a trait marker, trait-byillness interaction marker, or endophenotype of MDD that is independent of the acute depressive state and particularly active in a subgroup of MDD patients (Hasler et al., 2004; Iverson et al., 2011). We explored the potential molecular underpinnings of such a ‘cognitive’ MDD phenotype in a well-matched pilot sample of remitted MDD patients with poorer versus better performance on cognitive testing. Because the phenotype is likely highly complex, polygenic, and heterogeneous, we performed weighted gene coexpression network analysis (WGCNA), a hypothesis-free systems biology approach that identifies ‘modules’ of co-regulated – and therefore functionally related – genes in transcriptomic datasets. After constructing a gene co-expression network from samples of all patients, we determined whether modules were correlated with poorer versus better cognitive performance. We additionally explored whether modules were also correlated with a continuous measure of cognitive performance in both groups. We utilized whole blood transcriptomic data from remitted MDD patients from the Adelaide Cognitive Function and Mood Study (CoFaMS, HREC RAH protocol no. 111230g) matched for race, gender, age, Hamilton Depression Scale score, presence of psychotic features during any depressive episode, educational status, annual household income, current drugand alcohol use, and