Abstract:<h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Objective</h3><p>Despite there is a wide range of antidepressants available, with various mechanisms of actions, the efficacy of current therapeutic options is yet satisfactory. Previous shreds of evidence have indicated that genetics, cognitive, neuroendocrine, as well as personality factors, are all intrinsically linked and contribute to the diversity of treatment outcomes. We, therefore, sought to investigate this hypothesis in this study.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Method</h3><p>Based on 610 samples treated with a selection of serotonin reuptake inhibitor (SSRI), serotonin-norepinephrine reuptake inhibitors (SNRI), noradrenergic and specific serotonergic antidepressant (NaSSA) or tricyclic antidepressant (TCA), we compared the therapeutic effects of these four classes of drugs by survival analyses. Pharmacogenomic and survival analyses were carried out to explore the hereditary factors for curative effect and the accumulation of genetic factors was further discussed through pathway analysis and the global test. We built a machine learning-based prediction model that integrates genetic and non-genetic factors (including cognition, endocrinology, personality intelligence) to distinguish drug efficacy in single class drug situations. The values of the non-genetic makers after 6 weeks' treatment were collected to evaluate the efficacy of the model.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Results</h3><p>Our results from the 6-week antidepressant therapeutic study indicated that SSRI and SNRI are better treatments than those of TCA and NaSSA in the Chinese population. Among all possible paired single-agent survival analyses, citalopram and venlafaxine were more effective than mirtazapine. Allele C carriers at rs6354 (<em>SLC6A4</em>) and allele G carriers at rs12150214 (<em>SLC6A4</em>) were significantly prone to poorer treatment response to fluoxetine. Besides, the combination of three loci (rs929377-rs6191-rs32897) located in HPA pathway was significantly associated with the treatment outcome of fluoxetine. In female MDD patients, the minor allele of rs6323 and rs1137070 on the <em>MAOA</em> gene likely lead to a worse response to venlafaxine. Furthermore, genetic variants linked to drug efficacy tended to concentrate on the neurotrophin pathway in depressed patients comorbid with anxiety. From multivariate models, more severe cognitive deficits, psychopathic personality and lower levels of operational intelligence, and higher levels of cortisol predicted worse response status with SSRI or SNRI after 6-week treatment. Notably, genetic factors in the multi-dimensional prediction model for both classes of drugs include loci in <em>HTR2A</em> and <em>CRHBP</em> genes.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Conclusion</h3><p>SSRI and SNRI are more suitable for the treatment of Chinese people with depression. <em>SLC6A4</em> genetic variants, as well as HPA pathway, play an important role in the fluoxetine antidepressant therapeutic response while the polymorphism of <em>MAOA</em> gene involved in the pharmacological action of venlafaxine among female MDD patients. The presence of anxiety in MDD patients was related to the neurotrophin pathway. Genetic, cognitive, neuroendocrine, and personality intelligence factors combined have an ensemble impact on the medication effect of patients with major depression, leading to more precise and personalized medicine for specific groups of people.</p>

Personalized venlafaxine dose prediction using artificial intelligence technology: a retrospective analysis based on real-world data

Artificial intelligence-based model for dose prediction of sertraline in adolescents: a real-world study

Predicting quetiapine dose in patients with depression using machine learning techniques based on real-world evidence

Establishment of a predictive model for antidepressant efficacy based on machine learning

A Machine Learning Model For Predicting Blood Concentration of Quetiapine in Patients With Schizophrenia and Depression Based on Real‐World Data

Physiologically Based Pharmacokinetic Modeling to Unravel the Drug-gene Interactions of Venlafaxine: Based on Activity Score-dependent Metabolism by CYP2D6 and CYP2C19 Polymorphisms

Joint population pharmacokinetic modeling of venlafaxine and O-desmethyl venlafaxine in healthy volunteers and patients to evaluate the impact of morbidity and concomitant medication

AI-assisted prediction of differential response to antidepressant classes using electronic health records

Predicting plasma concentration of quetiapine in patients with depression using machine learning techniques based on real-world evidence

Using Machine Learning to Predict Antidepressant Treatment Outcome from Electronic Health Records

A machine learning approach to personalized dose adjustment of lamotrigine using noninvasive clinical parameters

Toward personalizing treatment for depression: predicting diagnosis and severity

PBPK modeling to predict the pharmacokinetics of venlafaxine and its active metabolite in different CYP2D6 genotypes and drug–drug interactions with clarithromycin and paroxetine

An individualized medication model of sodium valproate for patients with bipolar disorder based on machine learning and deep learning techniques

Predicting the effect of antidepressant treatment on relief from anxiety symptoms

Personalised treatment for cognitive impairment in dementia: development and validation of an artificial intelligence model

Influence and interaction of genetic, cognitive, neuroendocrine and personalistic markers to antidepressant response in Chinese patients with major depression

CYP2D6 Phenotype Influences Pharmacokinetic Parameters of Venlafaxine: Results from a Population Pharmacokinetic Model in Older Adults with Depression

The effectiveness of personalized medication based on drug-related genes, for schizophre-nia patients with resistance to traditional drugs.

Machine-Learning for Prescription Patterns: Random Forest in the Prediction of Dose and Number of Antipsychotics Prescribed to People with Schizophrenia

A Semi-parametric Estimation of Personalized Dose-response Function Using Instrumental Variables.