Least-absolute Shrinkage Operator Research Articles

A novel cancer-associated fibroblasts risk score model predict survival and immunotherapy in lung adenocarcinoma.

Lung adenocarcinoma (LUAD) is the leading cause of cancer-related death worldwide. Cancer-associated fibroblasts (CAFs) are a special type of fibroblasts, which play an important role in the development and immune escape of tumors. Weighted gene co-expression network analysis (WGCNA) was used to construct the co-expression module. In combination with univariate Cox regression and analysis of least absolute shrinkage operator (LASSO), characteristics associated with CAFs were developed for a prognostic model. The migration and proliferation of lung cancer cells were evaluated in vitro. Finally, the expression levels of proteins were analyzed by Western blot. LASSO Cox regression algorithm was then performed to select hub genes. Finally, a total of 2 Genes (COL5A2, COL6A2) were obtained. We then divided LUAD patients into high- and low-risk groups based on CAFs risk scores. Survival analysis, CAFs score correlation analysis and tumor mutation load analysis showed that COL5A2 and COL6A2 were high-risk genes for LUAD. Human Protein Atlas (HPA), western blot and PCR results showed that COL5A2 and COL6A2 were up-regulated in LUAD tissues. When COL5A2 and COL6A2 were knocked down, the proliferation, invasion and migration of lung cancer cells were significantly decreased. Finally, COL5A2 can affect LUAD progression through the Wnt/β-Catenin and TGF-β signaling pathways. Our CAFs risk score model offers a new approach for predicting the prognosis of LUAD patients. Furthermore, the identification of high-risk genes COL5A2 and COL6A2 and drug sensitivity analysis can provide valuable candidate clues for clinical treatment of LUAD.

Classifying Individuals With Rheumatic Conditions as Financially Insecure Using Electronic Health Record Data and Natural Language Processing: Algorithm Derivation and Validation.

We aimed to examine the feasibility of applying natural language processing (NLP) to unstructured electronic health record (EHR) documents to detect the presence of financial insecurity among patients with rheumatologic disease enrolled in an integrated care management program (iCMP). We incorporated supervised, rule-based NLP and statistical methods to identify financial insecurity among patients with rheumatic conditions enrolled in an iCMP (n = 20,395) in a multihospital EHR system. We constructed a lexicon for financial insecurity using data from available knowledge sources and then reviewed EHR notes from 538 randomly selected individuals (training cohort n = 366, validation cohort n = 172). We manually categorized records as having "definite," "possible," or "no" mention of financial insecurity. All available notes were processed using Narrative Information Linear Extraction, a rule-based version of NLP. Models were trained using the NLP features for financial insecurity using logistic, least absolute shrinkage operator (LASSO), and random forest performance characteristic and were compared with the reference standard. A total of 245,142 notes were processed from 538 individual patient records. Financial insecurity was present among 100 (27%) individuals in the training cohort and 63 (37%) in the validation cohort. The LASSO and random forest models performed identically and slightly better than logistic regression, with positive predictive values of 0.90, sensitivities of 0.29, and specificities of 0.98. The development of a context-driven lexicon used with rule-based NLP to extract data that identify financial insecurity is feasible for use and improved the capture for presence of financial insecurity with high accuracy. In the absence of a standard lexicon and construct definition for financial insecurity status, additional studies are needed to optimize the sensitivity of algorithms to categorize financial insecurity with construct validity.

Identification of AURKA as a Biomarker Associated with Cuproptosis and Ferroptosis in HNSCC.

Cuproptosis and ferroptosis represent copper- and iron-dependent forms of cell death, respectively, and both are known to play pivotal roles in head and neck squamous cell carcinoma (HNSCC). However, few studies have explored the prognostic signatures related to cuproptosis and ferroptosis in HNSCC. Our objective was to construct a prognostic model based on genes associated with cuproptosis and ferroptosis. We randomly assigned 502 HSNCC samples from The Cancer Genome Atlas (TCGA) into training and testing sets. Pearson correlation analysis was utilized to identify cuproptosis-associated ferroptosis genes in the training set. Cox proportional hazards (COX) regression and least absolute shrinkage operator (LASSO) were employed to construct the prognostic model. The performance of the prognostic model was internally validated using single-factor COX regression, multifactor COX regression, Kaplan-Meier analysis, principal component analysis (PCA), and receiver operating curve (ROC) analysis. Additionally, we obtained 97 samples from the Gene Expression Omnibus (GEO) database for external validation. The constructed model, based on 12 cuproptosis-associated ferroptosis genes, proved to be an independent predictor of HNSCC prognosis. Among these genes, the increased expression of aurora kinase A (AURKA) has been implicated in various cancers. To further investigate, we employed small interfering RNAs (siRNAs) to knock down AURKA expression and conducted functional experiments. The results demonstrated that AURKA knockdown significantly inhibited the proliferation and migration of HNSCC cells (Cal27 and CNE2). Therefore, AURKA may serve as a potential biomarker in HNSCC.

Integrative analysis reveals a four-gene signature for predicting survival and immunotherapy response in colon cancer patients using bulk and single-cell RNA-seq data.

Colon cancer (CC) ranks as one of the leading causes of cancer-related mortality globally. Single-cell transcriptome sequencing (scRNA-seq) offers precise gene expression data for distinct cell types. This study aimed to utilize scRNA-seq and bulk transcriptome sequencing (bulk RNA-seq) data from CC samples to develop a novel prognostic model. scRNA-seq data was downloaded from the GSE161277 database. R packages including "Seurat", "Harmony", and "singleR" were employed to categorize eight major cell types within normal and tumor tissues. By comparing tumor and normal samples, differentially expressed genes (DEGs) across these major cell types were identified. Gene Ontology (GO) enrichment analyses of DEGs for each cell type were conducted using "Metascape". DEGs-based signature construction involved Cox regression and least absolute shrinkage operator (LASSO) analyses, performed on The Cancer Genome Atlas (TCGA) training cohort. Validation occurred in the GSE39582 and GSE33382 datasets. The expression pattern of prognostic genes was verified using spatial transcriptome sequencing (ST-seq) data. Ultimately, an established prognostic nomogram based on the gene signature and age was established and calibrated. Sensitivity to chemotherapeutic drugs was predicted with the "oncoPredict" R package. Using scRNA-Seq data, we examined 33,213 cells, categorizing them into eight cell types within normal and tumor samples. GO enrichment analysis revealed various cancer-related pathways across DEGs in these cell types. Among the 55 DEGs identified via univariate Cox regression, four independent prognostic genes emerged: PTPN6, CXCL13, SPINK4, and NPDC1. Expression validation through ST-seq confirmed PTPN6 and CXCL13 predominance in immune cells, while SPINK4 and NPDC1 were relatively epithelial cell-specific. Creating a four-gene prognostic signature, Kaplan-Meier survival analyses emphasized higher risk scores correlating with unfavorable prognoses, confirmed across training and validation cohorts. The risk score emerged as an independent prognostic factor, supported by a reliable nomogram. Intriguingly, drug sensitivity analysis unveiled contrasting anti-cancer drug responses in the two risk groups, suggesting significant clinical implications. We developed a novel prognostic four-gene risk model, and these genes may act as potential therapeutic targets for CC.

Identification and validation of a novel cuproptosis-related stemness signature to predict prognosis and immune landscape in lung adenocarcinoma by integrating single-cell and bulk RNA-sequencing.

Cancer stem cells (CSCs) play vital roles in lung adenocarcinoma (LUAD) recurrence, metastasis, and drug resistance. Cuproptosis has provided a novel insight into the treatment of lung CSCs. However, there is a lack of knowledge regarding the cuproptosis-related genes combined with the stemness signature and their roles in the prognosis and immune landscape of LUAD. Cuproptosis-related stemness genes (CRSGs) were identified by integrating single-cell and bulk RNA-sequencing data in LUAD patients. Subsequently, cuproptosis-related stemness subtypes were classified using consensus clustering analysis, and a prognostic signature was constructed by univariate and least absolute shrinkage operator (LASSO) Cox regression. The association between signature with immune infiltration, immunotherapy, and stemness features was also investigated. Finally, the expression of CRSGs and the functional roles of target gene were validated in vitro. We identified six CRSGs that were mainly expressed in epithelial and myeloid cells. Three distinct cuproptosis-related stemness subtypes were identified and associated with the immune infiltration and immunotherapy response. Furthermore, a prognostic signature was constructed to predict the overall survival (OS) of LUAD patients based on eight differently expressed genes (DEGs) with cuproptosis-related stemness signature (KLF4, SCGB3A1, COL1A1, SPP1, C4BPA, TSPAN7, CAV2, and CTHRC1) and confirmed in validation cohorts. We also developed an accurate nomogram to improve clinical applicability. Patients in the high-risk group showed worse OS with lower levels of immune cell infiltration and higher stemness features. Ultimately, further cellular experiments were performed to verify the expression of CRSGs and prognostic DEGs and demonstrate that SPP1 could affect the proliferation, migration, and stemness of LUAD cells. This study developed a novel cuproptosis-related stemness signature that can be used to predict the prognosis and immune landscape of LUAD patients, and provided potential therapeutic targets for lung CSCs in the future.

Development of a Cancer-Associated Fibroblast-Related Prognostic Model in Breast Cancer via Bulk and Single-Cell RNA Sequencing.

The most numerous cells in the tumor microenvironment, cancer-associated fibroblasts (CAFs) play a crucial role in cancer development. Our objective was to develop a cancer-associated fibroblast breast cancer predictive model. We acquire breast cancer (BC) scRNA-seq data from Gene Expression Omnibus (GEO), and "Seurat" was used for data processing, including quality control, filtering, principal component analysis, and t-SNE. Afterward, "singleR" software was used to annotate cells. Seurat's "FindAllMarkers" program is used to locate particular CAF markers. clusterProfiler was used to analyze Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment. The Cancer Genome Atlas (TCGA) database was utilized to provide univariate Cox regression, least absolute shrinkage operator (LASSO) analysis using bulk RNA-seq data. For model development, multivariate Cox regression studies are used. Utilizing pRRophetic and Tumor Immune Dysfunction and Exclusion (TIDE) algorithms, chemosensitivity and immunotherapy response were predicted. The "rms" software was used to facilitate and simplify modeling. Integrating the scRNA-seq (GSE176078) dataset yielded 28 cell clusters. In addition, well-known cell types helped identify 12 cell types. We found 193 marker genes that are elevated in CAFs. In addition, a five-gene predictive model associated to CAF was created in the training set. In the training set, the validation set, and the external validation set, greater risk scores were associated with a worse prognosis. And individuals with a higher risk score were more susceptible to immunotherapy and conventional chemotherapy medicines. In conclusion, we establish a strong prognostic model comprised of 5 genes related with CAF that might serve as a potent prognostic indicator and aid clinicians in making more rational medication choices.

Cancer-associated fibroblast-related gene signatures predict survival and drug response in patients with colorectal cancer.

Background: Cancer-associated fibroblasts (CAFs) play an important role in the tumorigenesis, immunosuppression and metastasis of colorectal cancer (CRC), and can predict poor prognosis in patients with CRC. The present study aimed to construct a CAFs-related prognostic signature for CRC. Methods: The clinical information and corresponding RNA data of CRC patients were downloaded from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. The Estimation of STromal and Immune cells in MAlignant Tumor tissues (ESTIMATES) and xCell methods were applied to evaluate the tumor microenvironment infiltration from bulk gene expression data. Weighted gene co-expression network analysis (WGCNA) was used to construct co-expression modules. The key module was identified by calculating the module-trait correlations. The univariate Cox regression and least absolute shrinkage operator (LASSO) analyses were combined to develop a CAFs-related signature for the prognostic model. Moreover, pRRophetic and Tumor Immune Dysfunction and Exclusion (TIDE) algorithms were utilized to predict chemosensitivity and immunotherapy response. Human Protein Atlas (HPA) databases were employed to evaluate the protein expressions. Results: ESTIMATES and xCell analysis showed that high CAFs infiltration was associated with adverse prognoses. A twenty-gene CAFs-related prognostic signature (CAFPS) was established in the training cohort. Kaplan-Meier survival analyses reveled that CRC patients with higher CAFs risk scores were associated with poor prognosis in each cohort. Univariate and multivariate Cox regression analyses verified that CAFPS was as an independent prognostic factor in predicting overall survival, and a nomogram was built for clinical utility in predicting CRC prognosis. Patients with higher CAFs risk scores tended to not respond to immunotherapy, but were more sensitive to five conventional chemotherapeutic drugs. Conclusion: In summary, the CAFPS could serve as a robust prognostic indicator in CRC patients, which might help to optimize risk stratification and provide a new insight into individual treatments for CRC.

Identification of molecular subtypes and prognostic signature for hepatocellular carcinoma based on genes associated with homologous recombination deficiency

Hepatocellular carcinoma (HCC) is a rapidly developing digestive tract carcinoma. The prognosis of patients and side effects caused by clinical treatment should be better improved. Nonnegative matrix factorization (NMF) clustering was performed using 109 homologous recombination deficiency (HRD)-related of HCC genes from The Cancer Genome Atlas (TCGA) database. Limma was applied to analyze subtype differences. Immune scores and clinical characteristics of different subtypes were compared. An HRD signature were built with least absolute shrinkage operator (LASSO) and multivariate Cox analysis. Performance of the signature system was then assessed by Kaplan–Meier curves and receiver operating characteristic (ROC) curves. We identified two molecular subtypes (C1 and C2), with C2 showing a significantly better prognosis than C1. C1 contained 3623 differentially expressed genes. A 4-gene prognostic signature for HCC was established, and showed a high predicting accuracy in validation sets, entire TCGA data set, HCCDB18 and GSE14520 queues. Moreover, the risk score was validated as an independent prognostic marker for HCC. Our research identified two molecular subtypes of HCC, and proposed a novel scoring system for evaluating the prognosis of HCC in clinical practice.

Integrated single-cell and bulk RNA sequencing analysis identifies a cancer associated fibroblast-related signature for predicting prognosis and therapeutic responses in colorectal cancer

BackgroundCancer-associated fibroblasts (CAFs) contribute notably to colorectal cancer (CRC) tumorigenesis, stiffness, angiogenesis, immunosuppression and metastasis, and could serve as a promising therapeutic target. Our purpose was to construct CAF-related prognostic signature for CRC.MethodsWe performed bioinformatics analysis on single-cell transcriptome data derived from Gene Expression Omnibus (GEO) and identified 208 differentially expressed cell markers from fibroblasts cluster. Bulk gene expression data of CRC was obtained from The Cancer Genome Atlas (TCGA) and GEO databases. Univariate Cox regression and least absolute shrinkage operator (LASSO) analyses were performed on TCGA training cohort (n = 308) for model construction, and was validated in TCGA validation (n = 133), TCGA total (n = 441), GSE39582 (n = 470) and GSE17536 (n = 177) datasets. Microenvironment Cell Populations-counter (MCP-counter) and Estimate the Proportion of Immune and Cancer cells (EPIC) methods were applied to evaluated CAFs infiltrations from bulk gene expression data. Real-time polymerase chain reaction (qPCR) was performed in tissue microarrays containing 80 colon cancer samples to further validate the prognostic value of the CAF model. pRRophetic and Tumor Immune Dysfunction and Exclusion (TIDE) algorithms were utilized to predict chemosensitivity and immunotherapy response. Human Protein Atlas (HPA) databases and immunohistochemistry were used to evaluate the protein expressions.ResultsA nine-gene prognostic CAF-related signature was established in training cohort. Kaplan–Meier survival analyses revealed patients with higher CAF risk scores were correlated with adverse prognosis in each cohort. MCP-counter and EPIC results consistently revealed CAFs infiltrations were significantly higher in high CAF risk group. Patients with higher CAF risk scores were more prone to not respond to immunotherapy, but were more sensitive to several conventional chemotherapeutics, suggesting a potential strategy of combining chemotherapy with anti-CAF therapy to improve the efficacy of current T-cell based immunotherapies. Univariate and multivariate Cox regression analyses verified the CAF model was as an independent prognostic indicator in predicting overall survival, and a CAF-based nomogram was then built for clinical utility in predicting prognosis of CRC.ConclusionTo conclude, the CAF-related signature could serve as a robust prognostic indicator in CRC, which provides novel genomics evidence for anti-CAF immunotherapeutic strategies.

Characteristics of hypoxic tumor microenvironment in non-small cell lung cancer, involving molecular patterns and prognostic signature.

BackgroundThe mechanisms of hypoxia or immune microenvironment in cancer have been studied respectively, but the role of hypoxia immune microenvironment in non-small cell lung cancer (NSCLC) still needs further exploration.MethodsBy applying the K-means algorithm, 1,121 patients with NSCLC were divided into three categories. We evaluated the constructed signature in order to link it with the prognosis, which was constructed by univariate and least absolute shrinkage operator (LASSO) Cox regression analysis.ResultsA total of three clusters were obtained by clustering five Gene Expression Omnibus (GEO) data sets. Gene Set Variation Analysis (GSVA) and immune infiltration analysis were performed to explore the biological behavior. Cluster one presented an activated state of oncogenic pathways, and compared with the other two clusters, the median risk score was the highest, which was the reason for its poor survival. Cluster three showed that the immune pathway was active and the median risk score was the lowest, so the survival was the best. However, cluster two presented a state in which both immune and matrix pathways were activate. This was manifested as mutual antagonism, and its risk score was in the middle. Its survival was in the middle.ConclusionsThis work revealed the role of hypoxia related genes (HRGs) modification in tumor microenvironment, which was conducive to our comprehensive analysis of the prognosis of NSCLC, and provided direction and guidance for clinical immunotherapy.

Machine learning algorithm for characterizing risks of hypertension, at an early stage in Bangladesh

Background and aimsHypertension has become a major public health issue as the prevalence and risk of premature death and disability among adults due to hypertension has increased globally. The main objective is to characterize the risk factors of hypertension among adults in Bangladesh using machine learning (ML) algorithms. Materials and methodsThe hypertension data was derived from Bangladesh demographic and health survey, 2017–18, which included 6965 people aged 35 and above. Two most promising risk factor identification methods, namely least absolute shrinkage operator (LASSO) and support vector machine recursive feature elimination (SVMRFE) are implemented to detect the critical risk factors of hypertension. Additionally, four well-known ML algorithms as artificial neural network, decision tree, random forest, and gradient boosting (GB) have been used to predict hypertension. Performance scores of these algorithms were evaluated by accuracy, precision, recall, F-measure, and area under the curve (AUC). ResultsThe results clarify that age, BMI, wealth index, working status, and marital status for LASSO and age, BMI, marital status, diabetes and region for SVMRFE appear to be the top-most five significant risk factors for hypertension. Our findings reveal that the combination of SVMRFE-GB gives the maximum accuracy (66.98%), recall (97.92%), F-measure (78.99%), and AUC (0.669) compared to others. ConclusionGB-based algorithm confirms the best performer for prediction of hypertension, at an early stage in Bangladesh. Therefore, this study highly suggests that the policymakers make proper judgments for controlling hypertension using SVMRFE-GB-based combination to save time and reduce cost for Bangladeshi adults.

Do local and global factors impact the emerging markets' sovereign yield curves? Evidence from a data‐rich environment

AbstractThis paper investigates the relation between yield curve and macroeconomic factors for 10 emerging sovereign bond markets using the sample from January 2006 to April 2019. To this end, the diffusion indices obtained under four categories (global variables, inflation, domestic financial variables, and economic activity) are incorporated by estimating dynamic panel data regressions together with the yield curve factors. Besides, in order to capture dynamic interaction between yield curve and macroeconomic/financial factors, a panel vector autoregressive (VAR) analysis based on the system generalized method of moments (GMM) approach is utilized. Empirical results suggest that the level factor responds to shocks originated from inflation, domestic financial variables, and global variables. Furthermore, the slope factor is affected by shocks in global variables, and the curvature factor appears to be influenced by domestic financial variables. We also show that macroeconomic/financial factors captures significant predictive information over yield curve factors by running individual country factor‐augmented predictive regressions and variable selection algorithms such ridge regression, least absolute shrinkage operator (LASSO), and Elastic Net. Our findings have important implications for policymakers and fund managers by explaining the underlying forces of movements in the yield curve and forecasting accurately dynamics of yield curve factors.

Forecasting volatility using double shrinkage methods

In this paper, we propose and evaluate a shrinkage based methodology that is designed to improve the accuracy of volatility forecasts. Our approach is based on a two-step procedure for extracting latent common volatility factors from a large dimensional and high-frequency dataset. In the first step, we apply either least absolute shrinkage operator (LASSO) or the elastic net (EN) shrinkage on estimated integrated volatilities, in order to select a subset of assets that are informative about the target asset. In the second step, we utilize (sparse) principal component analysis on the selected assets, in order to estimate latent return factors, which are in turn used to construct latent volatility factors. Our two-step method is found to yield more accurate volatility predictions than a variety of alternative models based on approaches such as direct application of (S)PCA and direct application of LASSO or EN shrinkage, when comparing out-of-sample R2s and mean absolute forecasting errors, and when implementing predictive accuracy tests. Additionally model confidence sets are found to contain models solely based on our two-step approach. These forecasting gains are found to be robust to the use of original or log-scale realized volatility models, different data sampling frequencies, and different forecasting sub-periods.

Prognostic signature of lung adenocarcinoma based on stem cell-related genes

Lung adenocarcinoma (LUAD) is characterized by high infiltration and rapid growth. The function of the stem cell population is to control and maintain cell regeneration. Therefore, it is necessary to study the prognostic value of stem cell-related genes in LUAD. Signature genes were screened out from 166 stem cell-related genes according to the least absolute shrinkage operator (LASSO) and subsequently multivariate Cox regression analysis, and then established risk model. Immune infiltration and nomogram model were used to evaluate the clinical efficacy of signature. A signature consisting of 10 genes was used to dichotomize the LUAD patients into two groups (cutoff, 1.314), and then validated in GSE20319 and GSE42127. There was a significant correlation between signature and clinical characteristics. Patients with high-risk had a shorter overall survival. Furthermore, significant differences were found in multiple immune cells between the high-risk group and low-risk group. A high correlation was also reflected between signature and immune infiltration. What’s more, the signature could effectively predict the efficacy of chemotherapy in patients with LUAD, and a nomogram based on signature might accurately predict the prognosis of patients with LUAD. The signature-based of stem cell-related genes might be contributed to predicting prognosis of patients with LUAD.

Forecasting and nowcasting emerging market GDP growth rates: The role of latent global economic policy uncertainty and macroeconomic data surprise factors

AbstractIn this paper, we assess the predictive content of latent economic policy uncertainty and data surprise factors for forecasting and nowcasting gross domestic product (GDP) using factor‐type econometric models. Our analysis focuses on five emerging market economies: Brazil, Indonesia, Mexico, South Africa, and Turkey; and we carry out a forecasting horse race in which predictions from various different models are compared. These models may (or may not) contain latent uncertainty and surprise factors constructed using both local and global economic datasets. The set of models that we examine in our experiments includes both simple benchmark linear econometric models as well as dynamic factor models that are estimated using a variety of frequentist and Bayesian data shrinkage methods based on the least absolute shrinkage operator (LASSO). We find that the inclusion of our new uncertainty and surprise factors leads to superior predictions of GDP growth, particularly when these latent factors are constructed using Bayesian variants of the LASSO. Overall, our findings point to the importance of spillover effects from global uncertainty and data surprises, when predicting GDP growth in emerging market economies.

Forecasting Volatility Using Double Shrinkage Methods

Abstract In this paper, we propose and evaluate a shrinkage based methodology that is designed to improve the accuracy of volatility forecasts. Our approach is based on a two-step procedure for extracting latent common volatility factors from a large dimensional and high-frequency dataset. In the first step, we apply either least absolute shrinkage operator (LASSO) or the elastic net (EN) shrinkage on estimated integrated volatilities, in order to select a subset of assets that are informative about the target asset. In the second step, we utilize (sparse) principal component analysis on the selected assets, in order to estimate latent return factors, which are in turn used to construct latent volatility factors. Our two-step method is found to yield more accurate volatility predictions than a variety of alternative models based on approaches such as direct application of (S)PCA and direct application of LASSO or EN shrinkage, when comparing out-of-sample R 2 s and mean absolute forecasting errors, and when implementing predictive accuracy tests. Additionally model confidence sets are found to contain models solely based on our two-step approach. These forecasting gains are found to be robust to the use of original or log-scale realized volatility models, different data sampling frequencies, and different forecasting sub-periods.

Forecasting and Nowcasting Emerging Market GDP Growth Rates: The Role of Latent Global Economic Policy Uncertainty and Macroeconomic Data Surprise Factors

In this paper, we assess the predictive content of latent economic policy uncertainty and data surprises factors for forecasting and nowcasting GDP using factor-type econometric models. Our analysis focuses on five emerging market economies, including Brazil, Indonesia, Mexico, South Africa, and Turkey; and we carry out a forecasting horse-race in which predictions from various different models are compared. These models may (or may not) contain latent uncertainty and surprise factors constructed using both local and global economic datasets. The set of models that we examine in our experiments includes both simple benchmark linear econometric models as well as dynamic factor models (DFMs) that are estimated using a variety of frequentist and Bayesian data shrinkage methods based on the least absolute shrinkage operator (LASSO). We find that the inclusion of our new uncertainty and surprise factors leads to superior predictions of GDP growth, particularly when these latent factors are constructed using Bayesian variants of the LASSO. Overall, our findings point to the importance of spillover effects from global uncertainty and data surprises, when predicting GDP growth in emerging market economies.

ARCO: An Artificial Counterfactual Approach For High-Dimensional Panel Time-Series Data

We consider a new method to estimate causal effects when a treated unit suffers a shock or an intervention, such as a policy change, but there is not a readily available control group or counterfactual. We propose a two-step approach where in the first stage an artificial counterfactual is estimated from a large-dimensional set of variables from pool of untreated units (donors pool) using shrinkage methods, such as the Least Absolute Shrinkage Operator (LASSO). In the second stage, we estimate the average intervention effect on a vector of variables belonging to the treated unit, which is consistent and asymptotically normal. Our results are valid uniformly over a wide class of probability laws. Furthermore, we show that these results still hold when the date of the intervention is unknown and must be estimated from the data. Tests for multiple interventions and for contamination effects are also derived. By a simple transformation of the variables of interest, it is also possible to test for intervention effects on several moments (such as the mean or the variance) of the variables of interest. Finally, we can disentangle the actual intervention effects from confounding factors that usually bias before-and-after estimators. A detailed Monte Carlo experiment evaluates the properties of the method in finite samples and compares our proposal with other alternatives such as the differences-in-differences, factor models and the synthetic control method. An empirical application to evaluate the effects on inflation of a new anti tax evasion program in Brazil is considered. Our methodology is inspired by different branches of the literature such as: the Synthetic Control method, the Global Vector Autoregressive models, the econometrics of structural breaks, and the counterfactual analysis based on macro-econometric and panel data models.