Forecasting dengue and influenza incidences using a sparse representation of Google trends, electronic health records, and time series data

Dengue and influenza-like illness (ILI) are two of the leading causes of viral infection in the world and it is estimated that more than half the world’s population is at risk for developing these infections. It is therefore important to develop accurate methods for forecasting dengue and ILI incidences. Since data from multiple sources (such as dengue and ILI case counts, electronic health records and frequency of multiple internet search terms from Google Trends) can improve forecasts, standard time series analysis methods are inadequate to estimate all the parameter values from the limited amount of data available if we use multiple sources. In this paper, we use a computationally efficient implementation of the known variable selection method that we call the Autoregressive Likelihood Ratio (ARLR) method. This method combines sparse representation of time series data, electronic health records data (for ILI) and Google Trends data to forecast dengue and ILI incidences. This sparse representation method uses an algorithm that maximizes an appropriate likelihood ratio at every step. Using numerical experiments, we demonstrate that our method recovers the underlying sparse model much more accurately than the lasso method. We apply our method to dengue case count data from five countries/states: Brazil, Mexico, Singapore, Taiwan, and Thailand and to ILI case count data from the United States. Numerical experiments show that our method outperforms existing time series forecasting methods in forecasting the dengue and ILI case counts. In particular, our method gives a 18 percent forecast error reduction over a leading method that also uses data from multiple sources. It also performs better than other methods in predicting the peak value of the case count and the peak time.

Dengue and influenza-like illness (ILI) are leading causes of viral infection in the world and hence it is important to develop accurate methods for forecasting their incidence. We use Autoregressive Likelihood Ratio method, which is a computationally efficient implementation of the variable selection method, in order to obtain a sparse (non-lasso) representation of time series, Google Trends and electronic health records (for ILI) data. This method is used to forecast dengue incidence in five countries/states and ILI incidence in USA. We show that this method outperforms existing time series methods in forecasting these diseases. The method is general and can also be used to forecast other diseases.

Abstract P5-14-09: Integration of primary care practitioners in a breast multidisciplinary team meeting- A pilot study

Primary care practitioners are key to integrated patient care, as they often have an established relationship with patient and family, with knowledge of comorbidities and psychosocial factors. They can encourage compliance with cancer treatments and manage transition from acute care through survivorship.

Communication with primary care however has often been unsatisfactory, with information not helpful to that practitioner or not provided in a timely manner.

We recruited a general practitioner (GP) to attend the weekly breast cancer multidisciplinary meeting, with a responsibility to make contact with the primary care practitioner of patients presented prior to the meeting, so that relevant information from a primary care perspective could be taken into consideration for team deliberations. This also meant that the primary doctor was aware the patient was being discussed. Recommendations were fed back to that doctor after the meeting so that information was timely, and, being given by a peer, more useful.

The GP attended 12 consecutive MDTs where 105 patients were discussed. In 55% contact was able to be made with the patient's GP prior to the meeting. This was perceived as not usually changing management plans, but reinforced known medical and social concerns. Feedback to the patients GP was regarded favourably, with reported greater confidence in managing the patient, particularly those with complex needs, as well as improving hospital to community communication in general.

This pilot has shown that it is feasible to include primary care practitioners in a multidisciplinary meeting and improve interdisciplinary communication.

Citation Format: Pitcher ME, Hornung I, Rangarajan P. Integration of primary care practitioners in a breast multidisciplinary team meeting- A pilot study [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-14-09.

Estimation of Vector Autoregressive Parameters and Granger Causality From Noisy Multichannel Data

OBJECTIVE

The objective of this paper is to estimate the parameters of a multivariate autoregressive process from noisy multichannel data.

METHODS

Using a multivariate generalization of the Cadzow method, we propose a new method for estimating autoregressive parameters from noisy data: the nonlinear Cadzow method.

RESULTS

We show that our method outperforms existing multivariate methods such as higher order Yule-Walker method and Kalman EM method on simulated data. We apply our method to estimation of Granger causality from noisy data and again obtain superior results compared to previous methods. Finally, when applied to experimental local field potential data from monkey somatosensory and motor cortical areas, our method produces results consistent with cortical physiology.

CONCLUSION

The proposed nonlinear Cadzow method outperforms existing methods in obtaining denoised estimates of multivariate autoregressive parameters.

SIGNIFICANCE

Since multichannel recordings have become commonplace in biomedical applications ranging from discovering functional connectivity in the brain to speech data analysis and these recordings are inevitably contaminated by measurement noise, we believe our method has the potential for significant impact.

Transcriptional activation by hepatocyte nuclear factor-4 in a cell-free system derived from rat liver nuclei

Hepatocyte nuclear factor-4 (HNF4) regulates gene expression by binding to direct repeat motifs of the RG(G/T)TCA sequence separated by one nucleotide (DR1). In this study we demonstrate that endogenous HNF4 present in rat liver nuclear extracts, as well as purified recombinant HNF4, activates transcription from naked DNA templates containing multiple copies of the DR1 element linked to the adenovirus major late promoter. Recombinant HNF4 also activates transcription from the rat cellular retinol binding protein II (CRBPII) promoter in vitro. The region between -105 and -63 bp of this promoter is essential for HNF-mediated transactivation. The addition of a peptide containing the LXXLL motif abolished HNF4-mediated transactivation in vitro suggesting that LXXLL-containing protein factor(s) are involved in HNF4-mediated transactivation in rat liver nuclear extracts. This is the first report on transactivation by HNF4 in a cell-free system derived from rat liver nuclei.

Retinoic acid is a negative regulator of AP-1-responsive genes

We present evidence that retinoic acid can down-regulate transcriptional activation by the nuclear protooncogene c-jun. All three members of the retinoic acid receptor (RAR) subfamily (RAR alpha, RAR beta, and RAR gamma) can repress transcriptional induction of the human collagenase gene or a heterologous promoter that contains the collagenase promoter AP-1-binding site. In contrast, the retinoid X receptor fails to repress Jun/AP-1 activity, demonstrating a significant difference between the two regulatory systems through which retinoids exert their transcriptional control. Analysis of RAR alpha mutants in transfection studies reveals that the DNA-binding domain is important for the inhibition of Jun/AP-1 activity, even though the RAR does not bind the collagenase AP-1 site. Rather, gel-retardation assays reveal that bacterially expressed full-length RAR alpha inhibits binding of Jun protein to target DNA. These data suggest that the RAR alpha may form a nonproductive complex with c-Jun and provides a simple mechanisms by which retinoic acid may limit cell growth and possibly malignant progression.

Functional antagonism between oncoprotein c-Jun and the glucocorticoid receptor

We present evidence that the glucocorticoid receptor (GR) and transcription factor Jun/AP-1 can reciprocally repress one another's transcriptional activation by a novel mechanism that is independent of DNA binding. Overexpression of c-Jun prevents the glucocorticoid-induced activation of genes carrying a functional glucocorticoid response element (GRE). Conversely, GR is able to repress AP-1-mediated transcriptional activation. Mutant analysis reveals that the ligand binding and DNA binding domains of GR and the region including the leucine zipper of c-Jun are required for repression. Gel retardation analysis demonstrates that bacterially expressed c-Jun disrupts GR-GRE complexes. These data indicate that members of two distinct classes of transcription factors can oppose one another's activity through a mechanism likely involving protein-protein interactions.