Leave-one-out Research Articles

Using Key Predictors in an SVM Model for Differentiating Spinal Fractures and Herniated Intervertebral Discs in Preoperative Anesthesia Evaluation.

Background/Objectives: Spinal conditions, such as fractures and herniated intervertebral discs (HIVDs), are often challenging to diagnose due to overlapping clinical symptoms and the difficulty in assessing their functional impact. Accurate differentiation between these conditions is crucial for effective treatment, particularly in the context of preoperative anesthesia evaluation, where understanding the underlying condition can influence anesthesia planning and pain management. Methods and Materials: This study presents a Support Vector Machine (SVM) model designed to distinguish between spinal fractures and HIVDs using key clinical predictors, including age, gender, preoperative Visual Analog Scale (VAS) pain scores, and the number of spinal fractures. A retrospective analysis was conducted on a dataset of 199 patients diagnosed with these conditions. The SVM model, using a radial basis function (RBF) kernel, classified the conditions based on the selected predictors. Model performance was evaluated using precision, recall, accuracy, and the Kappa index, with Leave-One-Out (LOO) cross-validation applied to ensure robust results. Results: The SVM model achieved a precision of 92.1% for fracture cases and 91.2% for HIVDs, with recall rates of 98.1% for fractures and 70.5% for HIVDs. The overall accuracy was 92%, and the Kappa index was 0.76, indicating substantial agreement. The analysis revealed that age and VAS pain scores were the most critical predictors for accurately diagnosing these conditions. Conclusions: These results highlight the potential of the SVM model with an RBF kernel to reliably differentiate between spinal fractures and HIVDs using routine clinical data. Future work could enhance model performance by incorporating additional clinical parameters relevant to preoperative anesthesia evaluation.

Causal relationship between hypothyroidism and ulcerative colitis: a bidirectional Mendelian randomization study.

Ulcerative colitis (UC) and Hashimoto's thyroiditis frequently cooccur in patients with multiple autoimmune conditions, but the specific association between UC and hypothyroidism is unknown. We used Mendelian randomization (MR) methods to determine the causal relationship between UC and hypothyroidism. We obtained single nucleotide polymorphisms (SNPs) related to ulcerative colitis (UC) and hypothyroidism from genome-wide association studies (GWAS) available in the public database of the Integrated Epidemiology Unit (IEU). To assess the causal relationship between UC and hypothyroidism, we employed MR-Egger, weighted median, inverse variance weighted (IVW), simple mode, and weighted mode methods. Sensitivity analyses were performed using Cochran's Q test, the horizontal pleiotropy test, and the leave-one-out (LOO) method to assess the reliability of the MR data. The genes corresponding to instrumental variables (IVs) were subjected to Gene Ontology (GO) functional annotation, Kyoto Encyclopedia of the Genome (KEGG) pathway enrichment analysis, and protein-protein interaction (PPI) analysis to explore the mechanisms behind the causal relationships at the gene level. Forward MR analysis indicated that hypothyroidism was associated with an increased risk of UC (IVW: P = 0.02, OR = 9.71, 95% confidence interval (CI) = 1.36-69.46). In contrast, reverse MR did not demonstrate a causal relationship between UC and hypothyroidism (IVW: P = 0.53). Sensitivity analysis proved the reliability of the results. The PPI network revealed CD247, CD80, and STAT4 as central genes. GO and KEGG analyses revealed significant enrichment of the T cell, gamma interferon (IFN-γ), and programmed cell death-1 (PD-1)/programmed cell death ligand-1 (PD-L1) pathways. Hypothyroidism was a risk factor for UC. The balance of T-cell differentiation played an important role in the process of hypothyroidism-induced UC, and IL-21 might be the key to finding a cure. Enrichment of PD-1/PD-L1 might attenuate inflammation by suppressing the immune action of T cells.

Causal associations of MICB, CTSA, and MMP9 proteins with oral cancer: Mendelian randomization study

Oral cancer (ORCA) is the most prevalent histological subtype of oral malignancies in which immune modulation is relevant. The goal of this work was to employ Mendelian randomization (MR) to investigate the causal connection between the immune-related proteins MICB, CTSA, MMP9, and ORCA. The Open GWAS database of the Integrative Epidemiology Unit (IEU) was accessed to collect GWAS data for ORCA (ieu-b-4961), MICB (prot-a-1898), CTSA (prot-a-717) and MMP9 (prot-a-1921). From 372,373 samples, the ORCA dataset comprises 7,723,107 single nucleotide polymorphisms (SNPs). MICB, CTSA, and MMP9 all have 10,534,735 SNPs and 3,301 sample sizes. Then, the primary SVMR implementation approaches were weighted mode, simple mode, inverse variance weighted (IVW), weighted median, and MR-Egger. IVW was the most effective technique. A sensitivity study was also carried out to assess the correctness of SVMR data, with special focus devoted to heterogeneity, horizontal pleiotropy, and Leave-One-Out (LOO). MVMR was eventually implemented as well. A Mendelian randomization analysis of the three exposure factors in the dataset (ieu-b-94, ebi-a-GCST012237) was also performed to validate the study results. According to the SVMR results, there was a noteworthy causal interaction between ORCA and MICB (P = 0.0014), MMP9 (P = 0.0343), and CTSA (P = 0.0003). Furthermore, odds ratios (ORs) values revealed that MMP9 (OR = 1.0005) was an ORCA risk factor, whereas MICB (OR = 0.9994) and CTSA (OR = 0.9993) were security factors. The robustness of the SVMR findings was confirmed by the p-values of the heterogeneity and horizontal pleiotropy, both of which were greater than 0.05. The MVMR result did not affect any of the safety or hazard features of these three exposure factors. However, the P value for MMP9 was greater than 0.05, implying that MICB and CTSA may have a greater influence on ORCA than MMP9. The validation outcomes in both datasets harmonized with the findings from previous research, thereby solidifying the reliability of results. Our investigation provided a crucial resource for further research on the subject by demonstrating a causal relationship between ORCA and MICB, CTSA, and MMP9.

Challenges and Solutions for Leave-One-Out Biosensor Design in the Context of a Rugged Fitness Landscape

The leave-one-out (LOO) green fluorescent protein (GFP) approach to biosensor design combines computational protein design with split protein reconstitution. LOO-GFPs reversibly fold and gain fluorescence upon encountering the target peptide, which can be redefined by computational design of the LOO site. Such an approach can be used to create reusable biosensors for the early detection of emerging biological threats. Enlightening biophysical inferences for nine LOO-GFP biosensor libraries are presented, with target sequences from dengue, influenza, or HIV, replacing beta strands 7, 8, or 11. An initially low hit rate was traced to components of the energy function, manifesting in the over-rewarding of over-tight side chain packing. Also, screening by colony picking required a low library complexity, but designing a biosensor against a peptide of at least 12 residues requires a high-complexity library. This double-bind was solved using a “piecemeal” iterative design strategy. Also, designed LOO-GFPs fluoresced in the unbound state due to unwanted dimerization, but this was solved by fusing a fully functional prototype LOO-GFP to a fiber-forming protein, Drosophila ultrabithorax, creating a biosensor fiber. One influenza hemagglutinin biosensor is characterized here in detail, showing a shifted excitation/emission spectrum, a micromolar affinity for the target peptide, and an unexpected photo-switching ability.

First report on regression-based QSAR addressing pesticide dissipation half-life in plants: A step towards sustainable public health

The excessive use of pesticides (an important group of chemicals) in the agricultural as well as public sectors raises a health concern. Pesticides affect humans and other living organisms via the food chain. Therefore, it is very necessary to calculate the dissipation half-life of pesticides in plants. Experimental prediction of pesticide dissipation half-lives requires complex environmental conditions, high cost, and a long time. Thus, in-silico half-life predictions are suitable and the best alternative. Herein, a total of six PLS (partial least squares) models namely, M1 (overall), M2 (fruit), M3 (plant interior), M4 (leaf), M5 (plant surface), and M6 (whole plant) alongside two MLR (multiple linear regression) models i.e. M7 (fruit surface) and model M8 (straw) were generated using dissipation half-lives (log10(T1/2)) of pesticides in plants and their different parts. Models were constructed in strict accordance with the guidelines outlined by the Organization for Economic Co-operation and Development (OECD) and extensively validated using globally accepted validation metrics (determination coefficient (R2) = 0.610–0.795, leave-one-out (LOO) cross-validated correlation coefficient (Q2LOO) = 0.520–0.660, MAE-FITTED TRAIN (mean absolute error fitted train) = 0.119–0.148, MAE-LOOTRAIN = 0.132–0.177, predictive R2 or Q2F1 = 0.538–0.567, Q2F2 = 0.500–0.565, MAETEST = 0.122–0.232), confirming their accuracy, reliability, predictivity, and robustness. Lipophilicity, the presence of a cyclomatic ring, suphur, aromatic amine fragments, and chlorine atom fragments are responsible (+ve contribution) for high dissipation half-lives of pesticides in plants. In contrast, hydrophilicity, pyrazine fragments, and rotatable bonds reduce (−ve negative contribution) the dissipation half-lives of pesticides in plants. To address the real-world applicability, the models were employed to screen the PPDB (Pesticide Properties Database) database, which revealed the top 10 pesticides with the highest log(T1/2) in the whole plant and respective parts of the plant body. The present work will aid in developing safer and novel pesticides, regulatory risk assessment, various risk assessments for the sustenance of public health, screening of databases, and data-gap filling.

Elucidating the role of hepatic enzymes in spontaneous abortion: a Mendelian randomization approach.

While the hepatic enzymes Aspartate Aminotransferase (AST) and Alanine Aminotransferase (ALT) are crucial for liver function, their role in Spontaneous Abortion (SA) has not been thoroughly explored. Utilizing Mendelian Randomization (MR), this study aims to clarify the putative causal relationship between AST/ALT levels and SA. Genome-wide association study (GWAS) summary data for SA (finn-b-O15_ABORT_SPONTAN), AST (ukb-d-30650_raw), and ALT (ukb-d-30620_raw) were acquired from the Integrative Epidemiology Unit OpenGWAS database. Bidirectional MR analysis was conducted using MR-Egger, Weighted Median, Simple Mode, Weighted Mode, and Inverse Variance Weighted (IVW) algorithms, and the robustness of MR results was assessed through sensitivity analyses including Heterogeneity, Horizontal Pleiotropy, and Leave-One-Out (LOO) tests. The causal role of AST and ALT's coaction in SA was explored via multivariable MR (MVMR) analysis. The MR results via the IVW algorithm revealed a causal relation between both AST and ALT and SA (AST: P = 0.013; ALT: P = 0.017), identifying them as risk factors for SA (AST: odd ratio (OR) = 1.019; ALT: OR = 1.012). Sensitivity analysis substantiated the reliability of these results. Moreover, not notably causality was found between SA and AST/ALT (P > 0.05). Through MVMR analysis, AST and ALT demonstrated functional complementarity in the occurrence of SA, attributable to counterbalanced causalities (AST: P = 0.128; ALT: P = 0.899). The study substantiates a causal linkage between transaminase levels and SA, enhancing our understanding of their biological interaction and the regulatory mechanisms at play. These insights could have implications for identifying novel biomarkers and therapeutic targets for SA.

Causal relationship between Interleukin-27 expression levels and osteoporosis: a bidirectional mendelian randomization study

BackgroundThis study aimed to evaluate the causal relationship between Interleukin-27 (IL-27) and osteoporosis by bidirectional Mendelian randomization (MR) analysis.MethodsFirstly, the genome-wide association study summary data of osteoporosis (finn-b-M13_OSTEOPOROSIS) and IL-27 levels (ebi-a-GCST90012017) were picked out from the Integrative Epidemiology Unit (IEU) OpenGWAS database. After filtrating instrumental variables (IVs), the bidirectional MR analysis between IL-27 levels and osteoporosis was performed by MR-Egger, Weighted median, Simple mode, Weighted mode, and Inverse variance weighted (IVW). Subsequently, the sensitivity analysis was adopted to evaluate the reliability of the MR results via the Heterogeneity, Horizontal pleiotropy test and Leave-One-Out (LOO) analysis. Finally, the enrichment analysis of genes corresponding to SNPs related to IL-27 levels derived from eQTLGen database was executed to explore in depth the biological function and regulatory mechanism of these genes on osteoporosis occurrence.ResultsThe bidirectional MR results based on IVW method revealed that IL-27 level as a risk factor was causally related to osteoporosis (P = 0.004, odds ratio (OR) = 1.123, 95% confidence interval (CI) = 1.037–1.217), whereas osteoporosis was not in significant connection with IL-27 levels (P > 0.05). In regard to the sensitivity analysis for forward MR results, there was no heterogeneity and horizontal pleiotropy, and no SNPs relevant to IL-27 levels existed severe bias, suggesting the reliability of forward MR analysis. Furthermore, a total of 74 genes corresponding to 26 SNPs of IL-27 levels were obtained and were mainly involved in immune and inflammatory pathways including MyD88-dependent toll-like receptor signaling pathway, Toll-like receptor signaling pathway, cytosolic DNA-sensing pathway and so forth.ConclusionsThis study supported that IL-27 level as a risk factor was causally connected with osteoporosis and might regulate the disease occurrence and progression by means of immune and inflammatory mechanisms, which could provide important reference and evidence for further exploring the role of IL-27 in the development of osteoporosis.

Causality between migraine and cardiovascular disease: a bidirectional Mendelian randomization study

BackgroundWhile growing evidence suggests a relationship between migraine and cardiovascular disease, the genetic evidence for a causal relationship between migraine and cardiovascular disease is still scarce. Investigating the causal association between migraine and cardiovascular disease is vital.MethodsWe carried out a bidirectional Mendelian randomization (MR) study including discovery samples and replication samples using publicly available genome-wide association study (GWAS) summary datasets and stringent screening instrumental variables. Four different MR techniques—Inverse variance weighted (IVW), MR ‒Egger, weighted median, and weighted mode—as well as various sensitivity analyses—Cochran’s Q, IVW radial, leave-one-out (LOO), and MR-PRESSO—were utilized to investigate the causal relationship between cardiovascular disease and migraine.ResultsThe protective causal effects of genetically predicted migraine on coronary artery disease (OR, 0.881; 95% CI 0.790–0.982; p = 0.023) and ischemic stroke (OR, 0.912; 95% CI 0.854–0.974; p = 0.006) were detected in forward MR analysis but not in any other cardiovascular disease. Consistently, we also discovered protective causal effects of coronary atherosclerosis (OR, 0.865; 95% CI 0.797–0.940; p = 0.001) and myocardial infarction (OR, 0.798; 95% CI 0.668–0.952; p = 0.012) on migraine in reverse MR analysis.ConclusionWe found a potential protective effect of migraine on coronary artery disease and ischemic stroke and a potential protective effect of coronary atherosclerosis and myocardial infarction on migraine. We emphasised epidemiological and genetic differences and the need for long-term safety monitoring of migraine medications and future research to improve cardiovascular outcomes in migraine patients.

KNNDM CV: k-fold nearest-neighbour distance matching cross-validation for map accuracy estimation

Abstract. Random and spatial cross-validation (CV) methods are commonly used to evaluate machine-learning-based spatial prediction models, and the performance values obtained are often interpreted as map accuracy estimates. However, the appropriateness of such approaches is currently the subject of controversy. For the common case where no probability sample for validation purposes is available, in Milà et al. (2022) we proposed the nearest-neighbour distance matching (NNDM) leave-one-out (LOO) CV method. This method produces a distribution of geographical nearest-neighbour distances (NNDs) between test and training locations during CV that matches the distribution of NNDs between prediction and training locations. Hence, it creates predictive conditions during CV that are comparable to what is required when predicting a defined area. Although NNDM LOO CV produced largely reliable map accuracy estimates in our analysis, as a LOO-based method, it cannot be applied to the large datasets found in many studies. Here, we propose a novel k-fold CV strategy for map accuracy estimation inspired by the concepts of NNDM LOO CV: the k-fold NNDM (kNNDM) CV. The kNNDM algorithm tries to find a k-fold configuration such that the empirical cumulative distribution function (ECDF) of NNDs between test and training locations during CV is matched to the ECDF of NNDs between prediction and training locations. We tested kNNDM CV in a simulation study with different sampling distributions and compared it to other CV methods including NNDM LOO CV. We found that kNNDM CV performed similarly to NNDM LOO CV and produced reasonably reliable map accuracy estimates across sampling patterns. However, compared to NNDM LOO CV, kNNDM resulted in significantly reduced computation times. In an experiment using 4000 strongly clustered training points, kNNDM CV reduced the time spent on fold assignment and model training from 4.8 d to 1.2 min. Furthermore, we found a positive association between the quality of the match of the two ECDFs in kNNDM and the reliability of the map accuracy estimates. kNNDM provided the advantages of our original NNDM LOO CV strategy while bypassing its sample size limitations.

Associations between immune cell traits and autoimmune thyroid diseases: a bidirectional two-sample mendelian randomization study.

Autoimmune thyroid diseases (AITDs), mainly including Graves' disease (GD) and Hashimoto's thyroiditis (HT), are common autoimmune disorders characterized by abnormal immune responses targeting the thyroid gland. We conducted a bidirectional two-sample MR analysis using the largest dataset of peripheral immune cell phenotypes from Sardinia, and the AITD dataset from the 10th round of the FinnGen and the UK Biobank project. Instrumental variables (IVs) were rigorously selected based on the three assumptions of MR and analyzed using the Wald ratio, inverse-variance weighted (IVW), MR-Egger, and weighted median methods. Additionally, sensitivity analyses were performed using Cochrane's Q, the Egger intercept, the MR-PRESSO, and the leave-one-out (LOO) method to ensure the robustness of the results. The Steiger test was utilized to identify and exclude potential reverse causation. The results showed that 3, 3, and 11 immune cell phenotypes were significantly associated with the risk of AITD. In GD, the proportion of naive CD4-CD8- (DN) T cells in T cells and the proportion of terminally differentiated CD4+T cells in T cells showed the strongest inducing and protective effects, respectively. In HT, lymphocyte count and CD45 on CD4+T cells showed the strongest inducing and protective effects, respectively. In autoimmune hypothyroidism, CD127 CD8+T cell count and terminally differentiated DN T cell count exhibited the strongest inducing and protective effects, respectively. Through MR analysis, our study provides direct genetic evidence of the impact of immune cell traits on AITD risk and lays the groundwork for potential therapeutic and diagnostic target discovery.

Assessing the causal effects of Eubacterium and Rumphococcus on constipation: a Mendelian randomized study.

Constipation is affected by a number of risk variables, including cardiovascular disease and growth factors. However, the impacts of gut flora on constipation incidence has not been shown. This work, Single-Variable Mendelian Randomization (SVMR) was utilized to estimate the causal relationship between the Eubacterium genus or Rumphococcus, and constipation. Data for constipation, Eubacterium genus and Rumphococcus were taken from the Integrated Epidemiology Unit (IEU) open GWAS database. Including 218,792 constipation samples, and there were 16,380,466 Single Nucleotide Polymorphisms (SNPs) for constipation. The ids of Eubacterium genus and Rumphococcus were sourced from MiBioGen database. The sample count for the Eubacterium genus was 17,380, with 656 SNPs. In addition, the sample size for Rumphococcus was 15,339, with 545 SNPs. The SVMR was performed to assess the risk of Eubacterium genus and Rumphococcus in constipation using weighted median, MR Egger, simple mode, inverse variance weighted (IVW), and weighted mode. Finally, we did a sensitivity analysis that included a heterogeneity, horizontal pleiotropy, and Leave-One-Out (LOO) test to examine the viability of the MR data. The SVMR revealed that the Eubacterium genus and Rumphococcus were causally connected to constipation, with Rumphococcus (P = 0.042, OR = 1.074) as a hazardous factor and Eubacterium genus (P = 0.004, OR = 0.909) as a safety factor. Sensitivity tests then revealed the absence of variability between the constipation and the exposure factors (Eubacterium genus and Rumphococcus). Additionally, there were no other confounding factors and the examined SNPs could only influence constipation through the aforementioned exposure factors, respectively. As a result, the MR results were fairly robust. Our investigation verified the causal links between the Eubacterium genus or Rumphococcus, and constipation, with greater Rumphococcus expression increasing the likelihood of constipation and the opposite being true for the Eubacterium genus.

A mendelian randomization study investigating the causal relationships between 1400 serum metabolites and autoimmune diseases

ObjectiveThis study aims to explore the causal relationships between 1400 serum metabolites (SMs) and five autoimmune diseases (Myasthenia gravis [MG], Multiple sclerosis [MS], Systemic lupus erythematosus [SLE], Type 1 diabetes mellitus [T1DM], and Ulcerative colitis [UC]) through Mendelian randomization analysis. MethodData on MG, MS, SLE, T1DM, and UC were obtained from the IEU OpenGWAS Project database, while information on 1400 SMs was extracted from GWAS summary statistics provided by Chen et al. Causal relationships were assessed using the inverse variance weighted (IVW), MR-Egger, Weighted Median (WME), and Simple median (SME) methods. The robustness of instrumental variables was verified through computation of the F-statistic. Heterogeneity was evaluated using Cochran's Q test and the leave-one-out (LOO) method. Horizontal pleiotropy was assessed using MR-Egger regression and MR-PRESSO. ResultFollowing correction of the IVW P values using the False Discovery Rate (FDR) method, it was found that increased levels of 5-methyluridine (ribothymidine) (OR = 1.191, 95%CI 1.086–1.307, FDR-P = 0.000) and 2′-deoxyuridine (OR = 1.337, 95%CI 1.127–1.586, FDR-P = 0.001) were found to be correlated with a higher risk of MS. Conversely, the ratio of S-adenosylhomocysteine (SAH) to 5-methyluridine (ribothymidine) (OR = 0.771, 95%CI 0.649–0.916, FDR-P = 0.007) was linked to a decreased risk of MS. Levels of 1,2-dilinoleoyl-GPE (18:2/18:2) (OR = 0.877, 95%CI 0.791–0.974, FDR-P = 0.003) appear to be a protective factor for T1DM. No notable correlations between SMs and MG, SLE, or UC. The study detected no heterogeneity or horizontal pleiotropy. ConclusionLevels of 5-methyluridine (ribothymidine), 2′-deoxyuridine, and the ratio of S-adenosylhomocysteine (SAH) to 5-methyluridine (ribothymidine) can serve as predictors for MS. Similarly, 1,2-dilinoleoyl-GPE (18:2/18:2) levels can be used to predict T1DM. However, no significant causal relationships were found between SMs and MG, SLE, or UC. This observation holds significant clinical implications for crafting tailored preventive and therapeutic approaches for ADs.

Causal association between phenylalanine and Parkinson's disease: a two-sample bidirectional mendelian randomization study.

Research findings indicate a putative indirect or latent association between phenylalanine (Phe) and Parkinson's disease (PD). In this study, we aimed to analyze the causal relationship between Phe and PD by two sample Mendelian randomization (MR) analysis. In this study, the PD-related dataset and Phe-related dataset were downloaded from Integrative Epidemiology U1nit (IEU) Open Genome-Wide Association Study (GWAS) database. Four algorithms (MR Egger, maximum likelihood, inverse variance weighting (IVW) and unweighted regression) were used to perform MR analysis. The sensitivity analysis (heterogeneity test, horizontal pleiotropy test and Leave-One-Out (LOO) analysis) was used to assess the reliability of MR analyses. In the forward MR analysis, Phe was a safety factor for PD (p-value < 0.05 and odds ratios (OR) < 1). The results of reverse MR analysis showed that there was no causal relationship between PD and Phe (p-value > 0.05). In addition, sensitivity analysis showed that MR analysis was reliable. The results of this study revealed that Phe was a safety factor for PD, meaning that Phe reduced the risk of PD.

Quantitative Structure-Activity Relationship Study of Fatty Acid Derivatives as Lubricant Additive

Fatty acid compounds and their derivatives have been known in tribology as a source of bio-lubricant raw materials. New compounds were developed computationally to sort out the physical properties of the desired lubricant additive before synthesis in the laboratory. In this study, the correlation between the chemical structure of fatty acid derivative compounds and their physical properties through the quantitative structure-activity relationship (QSAR) using the density functional theory (DFT) was determined. QSAR model resulted in two linear regression equations for the wear rate performance test at (1) the value of the weld load with 5%, 500N variation and (2) the value of the wear scar diameter 5%, 150N variation. Validation of the regression model was conducted using the leave-one-out (LOO) method. The new additive lubricants were designed using the obtained QSAR models of the wear rate performance test with the descriptor of total energy, dipole moment, and softness. Stearoyl diethanolamine compound was proposed as a lubricant additive with the best performance parameter of weld load and wear scar diameter at 298.03 kg and 0.29 mm respectively. The percentage performance of stearoyl diethanolamine for weld load and wear scar diameter was increased by 24.18% and 47.60%, respectively.

Assessing the causal association of pregnancy complications with diabetes and cardiovascular disease.

To the best of our knowledge, numerous observational studies have linked pregnancy complications to increased risks of diabetes and cardiovascular disease (CVD), causal evidence remains lacking. Our aim was to estimate the association of adverse pregnancy outcomes with diabetes and cardiovascular diseases. A two-sample Mendelian randomization (MR) analysis was employed, which is not subject to potential reverse causality. Data for pregnancy complications were obtained from the FinnGen consortium. For primary analysis, outcome data on diabetes, related traits, stroke, and coronary heart disease (CHD) were extracted from the GWAS Catalog, MAGIC, MEGASTROKE, and CARDIoGRAMplusC4D consortium. The MAGIC and UKB consortium datasets were used for replication and meta-analysis. Causal effects were appraised using inverse variance weighted (IVW), weighted median (WM), and MR-Egger. Sensitivity analyses were implemented with Cochran's Q test, MR-Egger intercept test, MR-PRESSO, leave-one-out (LOO) analysis and the funnel plot. Genetically predicted gestational diabetes mellitus (GDM) was causally associated with an increased diabetes risk (OR=1.01, 95% CI=1-1.01, P<0.0001), yet correlated with lower 2-hour post-challenge glucose levels (OR=0.89, 95% CI=0.82-0.97, P=0.006). Genetic liability for pregnancy with abortive outcomes indicated decreased fasting insulin levels (OR=0.97, 95% CI=0.95-0.99, P=0.02), but potentially elevated glycated hemoglobin levels (OR=1.02, 95% CI=1.01-1.04, P=0.01). Additionally, hypertensive disorders in pregnancy was tentatively linked to increased risks of stroke (OR=1.11, 95% CI=1.04-1.18, P=0.002) and CHD (OR=1.3, 95% CI=1.2-1.4, P=3.11E-11). Gestational hypertension might have a potential causal association with CHD (OR=1.11, 95% CI=1.01-1.22, P=0.04). No causal associations were observed between preterm birth and diabetes, stroke, or CHD. The findings of this study provide genetic evidence that gestational diabetes, pregnancy with abortive outcomes, and hypertensive disorders in pregnancy may serve as early indicators for metabolic and cardiovascular risks. These insights are pivotal for the development of targeted screening and preventive strategies.

The causal effects of intelligence and fluid intelligence on Parkinson’s disease: a Mendelian randomization study

BackgroundParkinson’s disease (PD) is a chronic neurodegenerative disease that affects the central nervous system, primarily the motor nervous system, and occurs most often in older adults. A large number of studies have shown that high intelligence leads to an increased risk of PD. However, whether there is a causal relationship between intelligence on PD has not yet been reported.MethodsIn this study, Mendelian randomization (MR) analysis was performed with intelligence (ebi-a-GCST006250) and fluid intelligence score (ukb-b-5238) as exposure factors and PD (ieu-b-7) as an outcome, which the datasets were mined from the IEU OpenGWAS database. MR analysis was performed through 3 methods [MR Egger, weighted median, inverse variance weighted (IVW)], of which IVW was the primary method. In addition, the reliability of the results of the MR analysis was assessed via the heterogeneity test, the horizontal polytropy test, and Leave-One-Out (LOO). Finally, based on gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases, the genes corresponding to intelligence and fluid intelligence score related to SNPs were enriched for functional features and pathways.ResultsThe results of MR analysis suggested that elevated intelligence indicators can increase the risk of PD [p = 0.015, Odd Ratio (OR) = 1.316]. Meanwhile, fluid intelligence score was causally associated with the PD (p = 0.035), which was a risk factor (OR = 1.142). The reliability of the results of MR analysis was demonstrated by sensitivity analysis. Finally, the results of GO enrichment analysis for 87 genes corresponding to intelligence related SNPs mainly included regulation of synapse organization, developmental cell growth, etc. These genes were enriched in the synaptic vessel cycle, polycomb expressive complex in KEGG. Similarly, 44 genes corresponding to SNPs associated with fluid intelligence score were used for enrichment analysis. Based on the GO database, these genes were mainly enriched in regulation of developmental growth, negative regulation of neuron projection development, etc. In KEGG, 44 genes corresponding to SNPs associated with fluid intelligence score were enriched in signaling pathways including Alzheimer’s disease, the cellular senescence, etc.ConclusionThe causal relationships between intelligence and fluid intelligence scores, and PD were demonstrated through MR analysis, providing an important reference and evidence for the study of PD.

Machine learning allows for large-scale habitat prediction of a wide-ranging carnivore across diverse ecoregions

ContextResource selection functions are powerful tools for predicting habitat selection of animals. Recently, machine-learning methods such as random forest have gained popularity for predicting habitat selection due to their flexibility and strong predictive performance.ObjectivesWe tested two methods for predicting continental-scale, second-order habitat selection of a wide-ranging large carnivore, the mountain lion (Puma concolor), to support continent-wide conservation management, including estimating abundance, and to predict habitat suitability for recolonizing or reintroduced animals.MethodsWe compared a generalized linear model (GLM) and a random forest model using GPS location data from 476 individuals across 20 study sites in the western USA and Canada and remotely-sensed landscape data. We internally validated models and examined their ability to correctly classify used and available points by calculating area under the receiver operating characteristics (AUC). We performed leave-one-out (LOO) out-of-sample tests of predictive strength on both models.ResultsBoth models suggested that mountain lions select for steeper slopes, areas closer to water, and with higher normalized difference vegetation index (NDVI), and against variables associated with human impact. The random forest model (AUC = 0.94) demonstrated that mountain lion habitat can be accurately predicted at continental scales, outperforming the traditional GLM model (AUC = 0.68). Our LOO validation provided similar results (x̄ = 0.93 for the random forest and x̄ = 0.65 for the GLM).ConclusionsWe found that the added flexibility of the random forest model provided deeper insights into how individual covariates impacted habitat selection across diverse ecosystems. Our LOO analyses suggested that our model can predict mountain lion habitat selection in unoccupied areas or where local data are unavailable. Our model thus provides a tool to support discussions and analyses relevant to continent-wide mountain lion conservation and management including estimating metapopulation abundance.

Abstract PS06-07: Liquid biopsy determination of HER2 status in breast cancer: results from a novel epigenomic platform

Abstract Background: Tumor HER2 status remains a critical predictive biomarker for patients with breast cancer given the availability of highly effective anti-HER2 therapies. However, determination of HER2 status relies on a complex, tissue-dependent algorithm optimized for extremes of the phenotype – HER2+ vs HER2- status. Given the challenges in reliably obtaining informative tissue biopsies as disease progresses, the new therapeutic relevance of HER2-low breast cancer (IHC 1+ and FISH-negative IHC 2+), and the undependability of biomarkers in this range, better approaches are needed. Epigenomic signatures detectable in cell-free tumor DNA (cfDNA) may provide a minimally invasive, reliable determination of HER2 status. Here, we present a multimodal epigenomic liquid biopsy platform that could offer a minimally invasive alternative to tissue-based determination of transcriptionally regulated targets such as HER2. Methods: We identified patients with metastatic breast cancer (n=20) that were HER2-positive with IHC 3+ (n=6) or HER2-negative with IHC 0 (n=14) based on ASCO/CAP guidelines who underwent research biopsies at the Dana-Farber Cancer Institute from 2012-2023. Study participants had undergone concurrent blood sampling collected in Streck Tubes, processed as plasma, and stored at -80ºC. We applied a novel, multimodal epigenomic assay to profile tumor-derived gene regulatory programs from 1mL plasma. This genome-wide approach analyzed histone modifications associated with active gene promoters and enhancers, and DNA methylation. Classifier feature selection was derived from epigenomically profiled breast cancer cell lines. Final classifier performance was cross validated using a standard Leave one Out (LoO) scheme. Results: Our cross-validated classifier yielded an area under the ROC curve (AUC) of 0.91 in the 20 samples tested. When including only samples with detectable ctDNA as assessed by low pass whole genome sequencing (ichorCNA algorithm) (N=13), our classifier had an AUC of 0.98. Using regularized regression approach, we identified 33 loci with the highest predictive value; 8 loci were within the ERBB2 amplicon on chromosome 17, while the remaining 25 loci were in other genomic regions. We will present results from this and an expanded cohort, including participants with HER2 IHC 1+ and 2+ disease. Conclusions: We present a novel, minimally invasive HER2 classifier leveraging epigenomic information from cell-free tumor DNA from 1 mL plasma. If this approach remains robust in accurately classifying additional HER2 states, it could address the limitations of current tissue-based testing. Integration of this epigenomic platform in prospective interventional studies could support the development of novel and improved predictive classifiers for response to HER2-directed therapy. Citation Format: Heather Parsons, Sylvan Baca, Anthony D'Ippolito, Hyun-Hwan Jeong, Kalie Smith, Paolo Tarantino, Melissa Hughes, Kerry Sendrick, Corrie Painter, Ji-Heui Seo, Elizabeth Grant, Jamey Guess, Matthew Davidsohn, Gitanjali Lakshminarayanan, Sarah Strauss, Hunter Savignano, John Canniff, Brad Fortunato, Matthew Eaton, Toni Choueiri, Sara Tolaney, Matthew Freedman, Nancy Lin. Liquid biopsy determination of HER2 status in breast cancer: results from a novel epigenomic platform [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PS06-07.

Searching for homogeneous regions in regional flood frequency analysis for Southeast Australia

Study regionSoutheast Australia Study focusThis study examines identification of homogeneous regions for regional flood frequency analysis (RFFA) using 201 gauged catchments and applying principal component analysis (PCA) and cluster analysis. Quantile regression technique (QRT) is applied to develop prediction equations. Leave-one-out (LOO) validation is used to evaluate the accuracy of the developed prediction equations. New hydrologic insights for the regionFor the regions formed based on PCA and cluster analysis, the Hosking and Wallis’ heterogeneity (H1) statistics are found to vary in the range of 11.55–24.52, and 5.11–26.27, respectively. The assumed regions are highly heterogeneous as H1 values are much higher than 1.00. The median of absolute relative error values associated with the developed prediction equations for the assumed regions are found to vary in the range of 17–62 %, which are comparable to similar RFFA studies. PCA and cluster analysis do not deliver sufficiently homogeneous regions in southeast Australia. The degree of heterogeneity of a region has little effect on the relative accuracy of flood quantile estimates. The findings of this study will assist to update RFFA technique in Australian Rainfall and Runoff (national guideline).

Causal effects of childhood obesity on neuroticism and subjective well-being: A two-sample Mendelian randomization study

BackgroundChildhood obesity is linked to both neuroticism and subjective wellbeing (SWB); however, the causal relations between them remain unclear. MethodsTwo-sample Mendelian randomization (MR) analysis was applied to determine the causal effects of childhood BMI (n = 39,620) on neuroticism (n = 366,301) and SWB (n = 298,420) using summary statistics from large scale genome-wide association studies (GWASs). Inverse-variance weighting (IVW), weighted mode, weighted median, and MR-Egger approaches were used to estimate the causal effects. Sensitivity analyses including the Cochran's Q statistics, MR-Egger intercept test, MR-PRESSO global test, and the leave-one-out (LOO) analysis were used to assess potential heterogeneity and horizontal pleiotropy. Two-step MR mediation analysis was employed to explore the potential mediation effects of neuroticism on the causal relationship between childhood BMI and SWB. ResultsOur study revealed that genetically predicted higher childhood BMI was causally associated with increased neuroticism (beta = 0.045, 95%CI = 0.013,0.077, p = 6.066e-03) and reduced SWB (beta = −0.059, 95%CI = -0.093,-0.024, p = 9.585e-04). Sensitivity analyses didn't detect any significant heterogeneity and horizontal pleiotropy (all p > 0.05). Additionally, the two-step MR mediation analysis indicated that the causal relationship between childhood BMI and SWB was partially mediated by neuroticism (proportion of mediation effects in total effects: 21.3 %, 95%CI: 5.4 % to 37.2, p = 0.0088). ConclusionGenetically proxy for higher childhood BMI was associated with increased neuroticism and reduced SWB. Further studies were warranted to investigate the underlying molecular mechanisms and potential use of weight management for improving personality and SWB.