Agricultural insect pests as models for studying stress-induced evolutionary processes.

Agricultural insect pests (AIPs) are widely successful in adapting to natural and anthropogenic stressors, repeatedly overcoming population bottlenecks and acquiring resistance to intensive management practices. Although they have been largely overlooked in evolutionary studies, AIPs are ideal systems for understanding rapid adaptation under novel environmental conditions. Researchers have identified several genomic mechanisms that likely contribute to adaptive stress responses, including positive selection on de novo mutations, polygenic selection on standing allelic variation and phenotypic plasticity (e.g., hormesis). However, new theory suggests that stress itself may induce epigenetic modifications, which may confer heritable physiological changes (i.e., stress-resistant phenotypes). In this perspective, we discuss how environmental stress from agricultural management generates the epigenetic and genetic modifications that are associated with rapid adaptation in AIPs. We summarise existing evidence for stress-induced evolutionary processes in the context of insecticide resistance. Ultimately, we propose that studying AIPs offers new opportunities and resources for advancing our knowledge of stress-induced evolution.

Evolution of chemosensory genes in Colorado potato beetle, Leptinotarsa decemlineata.

Associating with plant hosts is thought to have elevated the diversification of insect herbivores, which comprise the majority of global species diversity. In particular, there is considerable interest in understanding the genetic changes that allow host-plant shifts to occur in pest insects and in determining what aspects of functional genomic diversity impact host-plant breadth. Insect chemoreceptors play a central role in mediating insect-plant interactions, as they directly influence plant detection and sensory stimuli during feeding. Although chemosensory genes evolve rapidly, it is unclear how they evolve in response to host shifts and host specialization. We investigate whether selection at chemosensory genes is linked to host-plant expansion from the buffalo burr, Solanum rostratum, to potato, Solanum tuberosum, in the super-pest Colorado potato beetle (CPB), Leptinotarsa decemlineata (Coleoptera: Chrysomelidae). First, to refine our knowledge of CPB chemosensory genes, we developed novel gene expression data for the antennae and maxillary-labial palps. We then examine patterns of selection at these loci within CPB, as well as compare whether rates of selection vary with respect to 9 closely related, non-pest Leptinotarsa species that vary in diet breadth. We find that rates of positive selection on olfactory receptors are higher in host-plant generalists, and this signal is particularly strong in CPB. These results provide strong candidates for further research on the genetic basis of variation in insect chemosensory performance and novel targets for pest control of a notorious super-pest.

Stabilization of Prebiotic Vesicles by Peptides Depends on Sequence and Chirality: A Mechanism for Selection of Protocell-Associated Peptides.

Cells require oligonucleotides and polypeptides with specific, homochiral sequences to perform essential functions, but it is unclear how such oligomers were selected from random sequences at the origin of life. Cells were probably preceded by simple compartments such as fatty acid vesicles, and oligomers that increased the stability, growth, or division of vesicles could have thereby increased in frequency. We therefore tested whether prebiotic peptides alter the stability or growth of vesicles composed of a prebiotic fatty acid. We find that three of 15 dipeptides tested reduce salt-induced flocculation of vesicles. All three contain leucine, and increasing their length increases the efficacy. Also, leucine-leucine but not alanine-alanine increases the size of vesicles grown by multiple additions of micelles. In a molecular simulation, leucine-leucine docks to the membrane, with the side chains inserted into the hydrophobic core of the bilayer, while alanine-alanine fails to dock. Finally, the heterochiral forms of leucine-leucine, at a high concentration, rapidly shrink the vesicles and make them leakier and less stable to high pH than the homochiral forms do. Thus, prebiotic peptide-membrane interactions influence the flocculation, growth, size, leakiness, and pH stability of prebiotic vesicles, with differential effects due to sequence, length, and chirality. These differences could lead to a population of vesicles enriched for peptides with beneficial sequence and chirality, beginning selection for the functional oligomers that underpin life.

Genome Resequencing Reveals Rapid, Repeated Evolution in the Colorado Potato Beetle.

Insecticide resistance and rapid pest evolution threatens food security and the development of sustainable agricultural practices, yet the evolutionary mechanisms that allow pests to rapidly adapt to control tactics remains unclear. Here, we examine how a global super-pest, the Colorado potato beetle (CPB), Leptinotarsa decemlineata, rapidly evolves resistance to insecticides. Using whole-genome resequencing and transcriptomic data focused on its ancestral and pest range in North America, we assess evidence for three, nonmutually exclusive models of rapid evolution: pervasive selection on novel mutations, rapid regulatory evolution, and repeated selection on standing genetic variation. Population genomic analysis demonstrates that CPB is geographically structured, even among recently established pest populations. Pest populations exhibit similar levels of nucleotide diversity, relative to nonpest populations, and show evidence of recent expansion. Genome scans provide clear signatures of repeated adaptation across CPB populations, with especially strong evidence of selection on insecticide resistance genes in different populations. Analyses of gene expression show that constitutive upregulation of candidate insecticide resistance genes drives distinctive population patterns. CPB evolves insecticide resistance repeatedly across agricultural regions, leveraging similar genetic pathways but different genes, demonstrating a polygenic trait architecture for insecticide resistance that can evolve from standing genetic variation. Despite expectations, we do not find support for strong selection on novel mutations, or rapid evolution from selection on regulatory genes. These results suggest that integrated pest management practices must mitigate the evolution of polygenic resistance phenotypes among local pest populations, in order to maintain the efficacy and sustainability of novel control techniques.

The role of structural variants in pest adaptation and genome evolution of the Colorado potato beetle, Leptinotarsa decemlineata (Say).

Structural variation has been associated with genetic diversity and adaptation. Despite these observations, it is not clear what their relative importance is for evolution, especially in rapidly adapting species. Here, we examine the significance of structural polymorphisms in pesticide resistance evolution of the agricultural super-pest, the Colorado potato beetle, Leptinotarsa decemlineata. By employing a parent offspring trio sequencing procedure, we develop highly contiguous reference genomes to characterize structural variation. These updated assemblies represent >100-fold improvement of contiguity and include derived pest and ancestral nonpest individuals. We identify >200,000 structural variations, which appear to be nonrandomly distributed across the genome as they co-occur with transposable elements and genes. Structural variations intersect with exons in a large proportion of gene annotations (~20%) that are associated with insecticide resistance (including cytochrome P450s), development, and transcription. To understand the role structural variations play in adaptation, we measure their allele frequencies among an additional 57 individuals using whole genome resequencing data, which represents pest and nonpest populations of North America. Incorporating multiple independent tests to detect the signature of natural selection using SNP data, we identify 14 genes that are probably under positive selection, include structural variations, and SNPs of elevated frequency within the pest lineages. Among these, three are associated with insecticide resistance based on previous research. One of these genes, CYP4g15, is coinduced during insecticide exposure with glycosyltransferase-13, which is a duplicated gene enclosed within a structural variant adjacent to the CYP4g15 genic region. These results demonstrate the significance of structural variations as a genomic feature to describe species history, genetic diversity, and adaptation.

Ripples at edges of blooming lilies and torn plastic sheets.

Ripples arise at edges of petals of blooming Lilium casablanca flowers and at edges of torn plastic sheets. In both systems, ripples are a consequence of excess length along the edge of a sheet. Through the use of time-lapse videos of blooming lilies and published images of torn plastic sheets, we find that ripples in both systems are well described by the scaling relationship a∝w(L-w), where a is amplitude, w is wavelength, and L is arc length. A phenomenological relationship previously reported for self-similar ripple patterns, namely ⟨a⟩∝⟨w⟩, can be recovered by assuming that buckling stress is constant. Excess length along petal edges can also influence their overall Gaussian curvature, such that petals invert from a cup shape to a saddle shape upon blooming. Previous simulations of these shape changes have assumed that petal thickness decreases at least quadratically. Here, we evaluate tomograms of several varieties of lily buds and find that this assumption is valid along the short axis of the buds, but not the long axis. A challenge of employing traditional tomography methods to measure petal thickness is that the sample is destroyed; a single bud cannot be followed through the entire blooming process. To address this challenge, we provide proof of principle that the nondestructive, label-free method of x-ray tomography produces high-contrast three-dimensional scans on time scales short enough to follow lily blooming.

Prebiotic Membranes and Micelles Do Not Inhibit Peptide Formation During Dehydration.

Cycles of dehydration and rehydration could have enabled formation of peptides and RNA in otherwise unfavorable conditions on the early Earth. Development of the first protocells would have hinged upon colocalization of these biopolymers with fatty acid membranes. Using atomic force microscopy, we find that a prebiotic fatty acid (decanoic acid) forms stacks of membranes after dehydration. Using LC-MS-MS (liquid chromatography-tandem mass spectrometry) with isotope internal standards, we measure the rate of formation of serine dipeptides. We find that dipeptides form during dehydration at moderate temperatures (55 °C) at least as fast in the presence of decanoic acid membranes as in the absence of membranes. Our results are consistent with the hypothesis that protocells could have formed within evaporating environments on the early Earth.

Prebiotic Protocell Membranes Retain Encapsulated Contents during Flocculation, and Phospholipids Preserve Encapsulation during Dehydration.

The first cell membranes were likely composed of single-chain amphiphiles such as fatty acids. An open question is whether fatty acid membranes could have functioned within evaporative lakes on the early Earth, which have been hypothesized to concentrate prebiotic reactants. Evaporation also concentrates monovalent salts, which in turn cause fatty acid membrane vesicles to flocculate; significant loss of encapsulated contents during flocculation would have impeded early cell evolution. Here, we tested whether fatty acid vesicles retain encapsulated contents after flocculation and after drying. We found that vesicles composed of 2:1 decanoic acid:decanol encapsulate calcein dye throughout a process of flocculation in saturated salt solution and subsequent disaggregation of vesicles by dilution of the salt. However, 30 minutes of complete dehydration disrupted encapsulation by fatty acid vesicles. In contrast, phospholipid vesicles maintained encapsulation. Our results reveal a selective pressure for protocells to incorporate phospholipids: while fatty acid membranes can retain encapsulated contents during periods of dilute and saturating salt, phospholipids are necessary for encapsulation during dry periods. Our results are consistent with the hypothesis that evaporative lakes were productive sites for prebiotic chemistry and the origin of cells.

Prebiotic Vesicles Retain Solutes and Grow by Micelle Addition after Brief Cooling below the Membrane Melting Temperature.

Replication of RNA genomes within membrane vesicles may have been a critical step in the development of protocells on the early Earth. Cold temperatures near 0 °C improve the stability of RNA and allow efficient copying, while some climate models suggest a cold early Earth, so the first protocells may have arisen in cold-temperature environments. However, at cold temperatures, saturated fatty acids, which would have been available on the early Earth, form gel-phase membranes that are rigid and restrict mobility within the bilayer. Two primary roles of protocell membranes are to encapsulate solutes and to grow by incorporating additional fatty acids from the environment. We test here whether fatty acid membranes in the gel phase accomplish these roles. We find that gel-phase membranes of 10-carbon amphiphiles near 0 °C encapsulate aqueous dye molecules as efficiently as fluid-phase membranes do, but the contents are released if the aqueous solution is frozen at -20 °C. Gel-phase membranes do not grow measurably by micelle addition, but growth resumes when membranes are warmed above the gel-liquid transition temperature. We find that longer, 12-carbon amphiphiles do not retain encapsulated contents near 0 °C. Together, our results suggest that protocells could have developed within environments that experience temporary cooling below the membrane melting temperature, and that membranes composed of relatively short-chain fatty acids would encapsulate solutes more efficiently as temperatures approached 0 °C.

Growth of Prebiotically Plausible Fatty Acid Vesicles Proceeds in the Presence of Prebiotic Amino Acids, Dipeptides, Sugars, and Nucleic Acid Components.

Fatty acid vesicles may have played a role in the origin of life as a major structural component of protocells, with the potential for encapsulation of genetic materials. Vesicles that grew and divided more rapidly than other vesicles could have had a selective advantage. Fatty acid vesicles grow by incorporating additional fatty acids from micelles, and certain prebiotic molecules (e.g., sugars, nucleobases, and amino acids) can bind to fatty acid vesicles and stabilize them. Here, we investigated whether the presence of a variety of biomolecules affects the overall growth of vesicles composed of decanoic acid, a prebiotically plausible fatty acid, upon micelle addition. We tested 31 molecules, including 15 dipeptides, 7 amino acids, 6 nucleobases or nucleosides, and 3 sugars. We find that the initial radius and final radius of vesicles are largely unaffected by the presence of the additional compounds. However, three dipeptides enhanced the initial rates of growth compared to control vesicles with no small molecules added; another three dipeptides decreased the initial rates of growth. We conclude that vesicles can indeed grow in the presence of a wide range of molecules likely to have been involved in the origin of life. These results imply that vesicles would have been able to grow in complex and heterogeneous chemical environments. We find that the molecules that enhance the initial growth rate tend to have hydrophobic groups (e.g., leucine), which may interact with the lipid membrane to affect growth rate; furthermore, the molecules that cause the largest decrease in initial growth rate are dipeptides containing a serine residue, which contains a hydroxyl group that could potentially hydrogen-bond with the fatty acid carboxylate groups.

Frontiers in Prebiotic Chemistry and Early Earth Environments.

The Prebiotic Chemistry and Early Earth Environments (PCE3) Consortium is a community of researchers seeking to understand the origins of life on Earth and in the universe. PCE3 is one of five Research Coordination Networks (RCNs) within NASA's Astrobiology Program. Here we report on the inaugural PCE3 workshop, intended to cross-pollinate, transfer information, promote cooperation, break down disciplinary barriers, identify new directions, and foster collaborations. This workshop, entitled, "Building a New Foundation", was designed to propagate current knowledge, identify possibilities for multidisciplinary collaboration, and ultimately define paths for future collaborations. Presentations addressed the likely conditions on early Earth in ways that could be incorporated into prebiotic chemistry experiments and conceptual models to improve their plausibility and accuracy. Additionally, the discussions that followed among workshop participants helped to identify within each subdiscipline particularly impactful new research directions. At its core, the foundational knowledge base presented in this workshop should underpin future workshops and enable collaborations that bridge the many disciplines that are part of PCE3.

Prebiotic amino acids bind to and stabilize prebiotic fatty acid membranes.

The membranes of the first protocells on the early Earth were likely self-assembled from fatty acids. A major challenge in understanding how protocells could have arisen and withstood changes in their environment is that fatty acid membranes are unstable in solutions containing high concentrations of salt (such as would have been prevalent in early oceans) or divalent cations (which would have been required for RNA catalysis). To test whether the inclusion of amino acids addresses this problem, we coupled direct techniques of cryoelectron microscopy and fluorescence microscopy with techniques of NMR spectroscopy, centrifuge filtration assays, and turbidity measurements. We find that a set of unmodified, prebiotic amino acids binds to prebiotic fatty acid membranes and that a subset stabilizes membranes in the presence of salt and Mg2+ Furthermore, we find that final concentrations of the amino acids need not be high to cause these effects; membrane stabilization persists after dilution as would have occurred during the rehydration of dried or partially dried pools. In addition to providing a means to stabilize protocell membranes, our results address the challenge of explaining how proteins could have become colocalized with membranes. Amino acids are the building blocks of proteins, and our results are consistent with a positive feedback loop in which amino acids bound to self-assembled fatty acid membranes, resulting in membrane stabilization and leading to more binding in turn. High local concentrations of molecular building blocks at the surface of fatty acid membranes may have aided the eventual formation of proteins.

Elevated rates of positive selection drive the evolution of pestiferousness in the Colorado potato beetle (Leptinotarsa decemlineata, Say).

Contextualizing evolutionary history and identifying genomic features of an insect that might contribute to its pest status is important in developing early detection and control tactics. In order to understand the evolution of pestiferousness, which we define as the accumulation of traits that contribute to an insect population's success in an agroecosystem, we tested the importance of known genomic properties associated with rapid adaptation in the Colorado potato beetle (CPB), Leptinotarsa decemlineata Say. Within the leaf beetle genus Leptinotarsa, only CPB, and a few populations therein, has risen to pest status on cultivated nightshades, Solanum. Using whole genomes from ten closely related Leptinotarsa species native to the United States, we reconstructed a high-quality species tree and used this phylogenetic framework to assess evolutionary patterns in four genomic features of rapid adaptation: standing genetic variation, gene family expansion and contraction, transposable element abundance and location, and positive selection at protein-coding genes. Throughout approximately 20 million years of history, Leptinotarsa species show little evidence of gene family turnover and transposable element variation. However, there is a clear pattern of CPB experiencing higher rates of positive selection on protein-coding genes. We determine that these rates are associated with greater standing genetic variation due to larger effective population size, which supports the theory that the demographic history contributes to rates of protein evolution. Furthermore, we identify a suite of coding genes under positive selection that are putatively associated with pestiferousness in the Colorado potato beetle lineage. They are involved in the biological processes of xenobiotic detoxification, chemosensation and hormone function.

Precision medicine for long-term depression outcomes using the Personalized Advantage Index approach: cognitive therapy or interpersonal psychotherapy?

BACKGROUND: Psychotherapies for depression are equally effective on average, but individual responses vary widely. Outcomes can be improved by optimizing treatment selection using multivariate prediction models. A promising approach is the Personalized Advantage Index (PAI) that predicts the optimal treatment for a given individual and the magnitude of the advantage. The current study aimed to extend the PAI to long-term depression outcomes after acute-phase psychotherapy. METHODS: Data come from a randomized trial comparing cognitive therapy (CT, n = 76) and interpersonal psychotherapy (IPT, n = 75) for major depressive disorder (MDD). Primary outcome was depression severity, as assessed by the BDI-II, during 17-month follow-up. First, predictors and moderators were selected from 38 pre-treatment variables using a two-step machine learning approach. Second, predictors and moderators were combined into a final model, from which PAI predictions were computed with cross-validation. Long-term PAI predictions were then compared to actual follow-up outcomes and post-treatment PAI predictions. RESULTS: One predictor (parental alcohol abuse) and two moderators (recent life events; childhood maltreatment) were identified. Individuals assigned to their PAI-indicated treatment had lower follow-up depression severity compared to those assigned to their PAI-non-indicated treatment. This difference was significant in two subsets of the overall sample: those whose PAI score was in the upper 60%, and those whose PAI indicated CT, irrespective of magnitude. Long-term predictions did not overlap substantially with predictions for acute benefit. CONCLUSIONS: If replicated, long-term PAI predictions could enhance precision medicine by selecting the optimal treatment for a given depressed individual over the long term.

The contribution of depressive 'disorder characteristics' to determinations of prognosis for adults with depression: an individual patient data meta-analysis.

BACKGROUND: This study aimed to investigate general factors associated with prognosis regardless of the type of treatment received, for adults with depression in primary care. METHODS: We searched Medline, Embase, PsycINFO and Cochrane Central (inception to 12/01/2020) for RCTs that included the most commonly used comprehensive measure of depressive and anxiety disorder symptoms and diagnoses, in primary care depression RCTs (the Revised Clinical Interview Schedule: CIS-R). Two-stage random-effects meta-analyses were conducted. RESULTS: Twelve (n = 6024) of thirteen eligible studies (n = 6175) provided individual patient data. There was a 31% (95%CI: 25 to 37) difference in depressive symptoms at 3-4 months per standard deviation increase in baseline depressive symptoms. Four additional factors: the duration of anxiety; duration of depression; comorbid panic disorder; and a history of antidepressant treatment were also independently associated with poorer prognosis. There was evidence that the difference in prognosis when these factors were combined could be of clinical importance. Adding these variables improved the amount of variance explained in 3-4 month depressive symptoms from 16% using depressive symptom severity alone to 27%. Risk of bias (assessed with QUIPS) was low in all studies and quality (assessed with GRADE) was high. Sensitivity analyses did not alter our conclusions. CONCLUSIONS: When adults seek treatment for depression clinicians should routinely assess for the duration of anxiety, duration of depression, comorbid panic disorder, and a history of antidepressant treatment alongside depressive symptom severity. This could provide clinicians and patients with useful and desired information to elucidate prognosis and aid the clinical management of depression.

Is social support pre-treatment associated with prognosis for adults with depression in primary care?

OBJECTIVE: Depressed patients rate social support as important for prognosis, but evidence for a prognostic effect is lacking. We aimed to test the association between social support and prognosis independent of treatment type, and the severity of depression, and other clinical features indicating a more severe illness. METHODS: Individual patient data were collated from all six eligible RCTs (n = 2858) of adults seeking treatment for depression in primary care. Participants were randomized to any treatment and completed the same baseline assessment of social support and clinical severity factors. Two-stage random effects meta-analyses were conducted. RESULTS: Social support was associated with prognosis independent of randomized treatment but effects were smaller when adjusting for depressive symptoms and durations of depression and anxiety, history of antidepressant treatment, and comorbid panic disorder: percentage decrease in depressive symptoms at 3-4 months per z-score increase in social support = -4.14(95%CI: -6.91 to -1.29). Those with a severe lack of social support had considerably worse prognoses than those with no lack of social support: increase in depressive symptoms at 3-4 months = 14.64%(4.25% to 26.06%). CONCLUSIONS: Overall, large differences in social support pre-treatment were associated with differences in prognostic outcomes. Adding the Social Support scale to clinical assessments may be informative, but after adjusting for routinely assessed clinical prognostic factors the differences in prognosis are unlikely to be of a clinically important magnitude. Future studies might investigate more intensive treatments and more regular clinical reviews to mitigate risks of poor prognosis for those reporting a severe lack of social support.

Personalized treatment selection in routine care: Integrating machine learning and statistical algorithms to recommend cognitive behavioral or psychodynamic therapy.

Objective: This study aims at developing a treatment selection algorithm using a combination of machine learning and statistical inference to recommend patients' optimal treatment based on their pre-treatment characteristics. Methods: A disorder-heterogeneous, naturalistic sample of N = 1,379 outpatients treated with either cognitive behavioral therapy or psychodynamic therapy was analyzed. Based on a combination of random forest and linear regression, differential treatment response was modeled in the training data (n = 966) to indicate each individual's optimal treatment. A separate holdout dataset (n = 413) was used to evaluate personalized recommendations. Results: The difference in outcomes between patients treated with their optimal vs. non-optimal treatment was significant in the training data, but non-significant in the holdout data (b = -0.043, p = .280). However, for the 50% of patients with the largest predicted benefit of receiving their optimal treatment, the average percentage of change on the BSI in the holdout data was 52.6% for their optimal and 38.4% for their non-optimal treatment (p = .017; d = 0.33 [0.06, 0.61]). Conclusion: A treatment selection algorithm based on a combination of ML and statistical inference might improve treatment outcome for some, but not all outpatients and could support therapists' clinical decision-making.

A demonstration of a multi-method variable selection approach for treatment selection: Recommending cognitive-behavioral versus psychodynamic therapy for mild to moderate adult depression.

Objective: We use a new variable selection procedure for treatment selection which generates treatment recommendations based on pre-treatment characteristics for adults with mild-to-moderate depression deciding between cognitive behavioral (CBT) versus psychodynamic therapy (PDT). Method: Data are drawn from a randomized comparison of CBT versus PDT for depression (N = 167, 71% female, mean-age = 39.6). The approach combines four different statistical techniques to identify patient characteristics associated consistently with differential treatment response. Variables are combined to generate predictions indicating each individual's optimal-treatment. The average outcomes for patients who received their indicated treatment versus those who did not were compared retrospectively to estimate model utility. Results: Of 49 predictors examined, depression severity, anxiety sensitivity, extraversion, and psychological treatment-needs were included in the final model. The average post-treatment Hamilton-Depression-Rating-Scale score was 1.6 points lower (95%CI = [0.5:2.8]; d = 0.21) for those who received their indicated-treatment compared to non-indicated. Among the 60% of patients with the strongest treatment recommendations, that advantage grew to 2.6 (95%CI = [1.4:3.7]; d = 0.37). Conclusions: Variable selection procedures differ in their characterization of the importance of predictive variables. Attending to consistently-indicated predictors may be sensible when constructing treatment selection models. The small N and lack of separate validation sample indicate a need for prospective tests before this model is used.

Personalized prediction of antidepressant v. placebo response: evidence from the EMBARC study.

BACKGROUND: Major depressive disorder (MDD) is a highly heterogeneous condition in terms of symptom presentation and, likely, underlying pathophysiology. Accordingly, it is possible that only certain individuals with MDD are well-suited to antidepressants. A potentially fruitful approach to parsing this heterogeneity is to focus on promising endophenotypes of depression, such as neuroticism, anhedonia, and cognitive control deficits. METHODS: Within an 8-week multisite trial of sertraline v. placebo for depressed adults (n = 216), we examined whether the combination of machine learning with a Personalized Advantage Index (PAI) can generate individualized treatment recommendations on the basis of endophenotype profiles coupled with clinical and demographic characteristics. RESULTS: Five pre-treatment variables moderated treatment response. Higher depression severity and neuroticism, older age, less impairment in cognitive control, and being employed were each associated with better outcomes to sertraline than placebo. Across 1000 iterations of a 10-fold cross-validation, the PAI model predicted that 31% of the sample would exhibit a clinically meaningful advantage [post-treatment Hamilton Rating Scale for Depression (HRSD) difference ⩾3] with sertraline relative to placebo. Although there were no overall outcome differences between treatment groups (d = 0.15), those identified as optimally suited to sertraline at pre-treatment had better week 8 HRSD scores if randomized to sertraline (10.7) than placebo (14.7) (d = 0.58). CONCLUSIONS: A subset of MDD patients optimally suited to sertraline can be identified on the basis of pre-treatment characteristics. This model must be tested prospectively before it can be used to inform treatment selection. However, findings demonstrate the potential to improve individual outcomes through algorithm-guided treatment recommendations.

Predictive modeling for response to lithium and quetiapine in bipolar disorder.

OBJECTIVES: Lithium and quetiapine are known to be effective treatments for bipolar disorder. However, little information is available to inform prediction of response to these medications. Machine-learning methods can identify predictors of response by examining variables simultaneously. Further evaluation of models on a test sample can estimate how well these models would generalize to other samples. METHODS: Data (N = 482) were drawn from a randomized clinical trial of outpatients with bipolar I or II disorder who received adjunctive personalized treatment plus either lithium or quetiapine. Elastic net regularization (ENR) was used to generate models for lithium and quetiapine; these models were evaluated on a test set. RESULTS: Predictions from the lithium model explained 17.4% of the variance in actual observed scores of patients who received lithium in the test set, while predictions from the quetiapine model explained 32.1% of the variance of patients that received quetiapine. Of the baseline variables selected, those with the largest parameter estimates were: severity of mania; attention-deficit/hyperactivity disorder (ADHD) comorbidity; nonsuicidal self-injurious behavior; employment; and comorbidity with each of two anxiety disorders (social phobia/society anxiety and agoraphobia). Predictive accuracy of the ENR model outperformed the simple and basic theoretical models. CONCLUSION: ENR is an effective approach for building optimal and generalizable models. Variables identified through this methodology can inform future research on predictors of response to lithium and quetiapine, as well as future modeling efforts of treatment choice in bipolar disorder.