From data to discovery: AI-guided analysis of disease-relevant molecules in spinal muscular atrophy (SMA).

Spinal Muscular Atrophy is caused by partial loss of survival of motoneuron (SMN) protein expression. The numerous interaction partners and mechanisms influenced by SMN loss result in a complex disease. Current treatments restore SMN protein levels to a certain extent, but do not cure all symptoms. The prolonged survival of patients creates an increasing need for a better understanding of SMA. Although many SMN-protein interactions, dysregulated pathways, and organ phenotypes are known, the connections among them remain largely unexplored. Monogenic diseases are ideal examples for the exploration of cause-and-effect relationships to create a network describing the disease-context. Machine learning tools can utilize such knowledge to analyze similarities between disease-relevant molecules and molecules not described in the disease so far. We used an artificial intelligence-based algorithm to predict new genes of interest. The transcriptional regulation of 8 out of 13 molecules selected from the predicted set were successfully validated in an SMA mouse model. This bioinformatic approach, using the given experimental knowledge for relevance predictions, enhances efficient targeted research in SMA and potentially in other disease settings.

Mining hidden knowledge: embedding models of cause-effect relationships curated from the biomedical literature.

Motivation: We explore the use of literature-curated signed causal gene expression and gene-function relationships to construct unsupervised embeddings of genes, biological functions and diseases. Our goal is to prioritize and predict activating and inhibiting functional associations of genes and to discover hidden relationships between functions. As an application, we are particularly interested in the automatic construction of networks that capture relevant biology in a given disease context. Results: We evaluated several unsupervised gene embedding models leveraging literature-curated signed causal gene expression findings. Using linear regression, we show that, based on these gene embeddings, gene-function relationships can be predicted with about 95% precision for the highest scoring genes. Function embedding vectors, derived from parameters of the linear regression model, allow inference of relationships between different functions or diseases. We show for several diseases that gene and function embeddings can be used to recover key drivers of pathogenesis, as well as underlying cellular and physiological processes. These results are presented as disease-centric networks of genes and functions. To illustrate the applicability of our approach to other machine learning tasks, we also computed embeddings for drug molecules, which were then tested using a simple neural network to predict drug-disease associations. Availability and implementation: Python implementations of the gene and function embedding algorithms operating on a subset of our literature-curated content as well as other code used for this paper are made available as part of the Supplementary data. Supplementary information: Supplementary data are available at Bioinformatics Advances online.

The Coronavirus Network Explorer: mining a large-scale knowledge graph for effects of SARS-CoV-2 on host cell function.

BACKGROUND: Leveraging previously identified viral interactions with human host proteins, we apply a machine learning-based approach to connect SARS-CoV-2 viral proteins to relevant host biological functions, diseases, and pathways in a large-scale knowledge graph derived from the biomedical literature. Our goal is to explore how SARS-CoV-2 could interfere with various host cell functions, and to identify drug targets amongst the host genes that could potentially be modulated against COVID-19 by repurposing existing drugs. The machine learning model employed here involves gene embeddings that leverage causal gene expression signatures curated from literature. In contrast to other network-based approaches for drug repurposing, our approach explicitly takes the direction of effects into account, distinguishing between activation and inhibition. RESULTS: We have constructed 70 networks connecting SARS-CoV-2 viral proteins to various biological functions, diseases, and pathways reflecting viral biology, clinical observations, and co-morbidities in the context of COVID-19. Results are presented in the form of interactive network visualizations through a web interface, the Coronavirus Network Explorer (CNE), that allows exploration of underlying experimental evidence. We find that existing drugs targeting genes in those networks are strongly enriched in the set of drugs that are already in clinical trials against COVID-19. CONCLUSIONS: The approach presented here can identify biologically plausible hypotheses for COVID-19 pathogenesis, explicitly connected to the immunological, virological and pathological observations seen in SARS-CoV-2 infected patients. The discovery of repurposable drugs is driven by prior knowledge of relevant functional endpoints that reflect known viral biology or clinical observations, therefore suggesting potential mechanisms of action. We believe that the CNE offers relevant insights that go beyond more conventional network approaches, and can be a valuable tool for drug repurposing. The CNE is available at https://digitalinsights.qiagen.com/coronavirus-network-explorer .

Bidirectional Text Messaging to Improve Adherence to Recommended Lipid Testing.

BACKGROUND: Synergies between technology and health care in the United States are accelerating, increasing opportunities to leverage these technologies to improve patient care. METHODS: This study was a collaboration between an academic study team, a rural primary care clinic, and a local nonprofit informatics company developing tools to improve patient care through population management. Our team created a text messaging management tool, then developed methods for and tested the feasibility of bidirectional text messaging to remind eligible patients about the need for lipid testing. We measured patient response to the text messages, then interviewed 8 patients to explore their text messaging experience. RESULTS: Of the 129 patients the clinic was able to contact by phone, 29.4% had no cell phone or text-messaging capabilities. An additional 20% refused to participate. Two thirds of the 28 patients who participated in the text messaging intervention (67.9%) responded to at least 1 of the up to 3 messages. Seven of 8 interviewed patients had a positive text-messaging experience. CONCLUSIONS: Bidirectional text messaging is a feasible and largely acceptable form of communication for test reminders that has the potential to reach large numbers of patients in clinical care.

Coronary heart disease risk in volunteer firefighters in Victoria, Australia.

Coronary heart disease (CHD) is a major problem for firefighters, even when considering the healthy worker effect (HWE). Although volunteer firefighters outnumber paid personnel, previous research has focused on paid US firefighters. By contrast, no CHD data for Australian firefighters exist. Risk factor data were collected from 2,943 Australian volunteer firefighters and CHD risk was compared with reference "low-risk" and Australian population data. Predicted CHD risk for male and female firefighter was 19.2% and 5.1%, respectively. Female firefighters high blood pressure and fasting glucose was significantly lower than the general population, whereas all other risk factors was similar to the general population. Firefighters' CHD risk was greater than other volunteer and paid emergency services, but the prevalence for most risk factors was similar to the general population. Therefore, Australian volunteer firefighters may not benefit from the HWE.

Causal analysis approaches in Ingenuity Pathway Analysis.

MOTIVATION: Prior biological knowledge greatly facilitates the meaningful interpretation of gene-expression data. Causal networks constructed from individual relationships curated from the literature are particularly suited for this task, since they create mechanistic hypotheses that explain the expression changes observed in datasets. RESULTS: We present and discuss a suite of algorithms and tools for inferring and scoring regulator networks upstream of gene-expression data based on a large-scale causal network derived from the Ingenuity Knowledge Base. We extend the method to predict downstream effects on biological functions and diseases and demonstrate the validity of our approach by applying it to example datasets. AVAILABILITY: The causal analytics tools 'Upstream Regulator Analysis', 'Mechanistic Networks', 'Causal Network Analysis' and 'Downstream Effects Analysis' are implemented and available within Ingenuity Pathway Analysis (IPA, http://www.ingenuity.com). SUPPLEMENTARY INFORMATION: Supplementary material is available at Bioinformatics online.

TFInfer: a tool for probabilistic inference of transcription factor activities.

SUMMARY: TFInfer is a novel open access, standalone tool for genome-wide inference of transcription factor activities from gene expression data. Based on an earlier MATLAB version, the software has now been extended in a number of ways. It has been significantly optimised in terms of performance, and it was given novel functionality, by allowing the user to model both time series and data from multiple independent conditions. With a full documentation and intuitive graphical user interface, together with an in-built data base of yeast and Escherichia coli transcription factors, the software does not require any mathematical or computational expertise to be used effectively. AVAILABILITY: http://homepages.inf.ed.ac.uk/gsanguin/TFInfer.html CONTACT: gsanguin@staffmail.ed.ac.uk SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The consistent delivery of negative pressure to wounds using reticulated, open cell foam and regulated pressure feedback.

Negative pressure wound therapy (NPWT) is used to manage wounds and promote wound healing. The most common form of NPWT utilizes reticulated, open cell foam (ROCF). Pressure is transferred to the wound by ROCF using T.R.A.C.™ Technology (regulated pressure feedback [RPF]) creating an environment that promotes healing. This study examines the effectiveness of ROCF versus gauze in inducing macrostrain and investigates the ability of NPWT/ROCF/RPF to consistently deliver negative pressure to the wound, compensating for constantly changing wound fluid characteristics. In an in-vitro model, ROCF induced significantly greater macrostrain than gauze demonstrating a 57% decrease in dressing surface area following negative pressure application. The decrease measured with gauze under suction (GUS) was insignificant. The NPWT/ROCF/RPF system consistently delivered negative pressure to the wound when compared to GUS or ROCF without RPF. Further, with the negative pressure source elevated 36 in (90 cm) above surrogate wounds, GUS demonstrated a 7- to 10-fold pressure drop when compared to NPWT/ROCF/RPF. Systems without RPF are limited because they cannot sense or measure pressure delivered at the wound. In situations where pressure drop occurs, neither the clinician nor patient would necessarily know that suboptimal pressure was being delivered to the wound. Therefore, a system with ROCF and RPF capability that effectively monitors and maintains the NPWT environment plays a crucial role in the optimal induction of macrostrain and microstrain. The ability of the NPWT/ROCF/RPF system to monitor and maintain controlled, consistent delivery of negative pressure would seem important to achieve desired clinical outcomes.

Accelerated preclinical testing using transplanted tumors from genetically engineered mouse breast cancer models.

PURPOSE: The use of genetically engineered mouse (GEM) models for preclinical testing of anticancer therapies is hampered by variable tumor latency, incomplete penetrance, and complicated breeding schemes. Here, we describe and validate a transplantation strategy that circumvents some of these difficulties. EXPERIMENTAL DESIGN: Tumor fragments from tumor-bearing MMTV-PyMT or cell suspensions from MMTV-PyMT, -Her2/neu, -wnt1, -wnt1/p53(+/-), BRCA1/p53(+/-), and C3(1)T-Ag mice were transplanted into the mammary fat pad or s.c. into naïve syngeneic or immunosuppressed mice. Tumor development was monitored and tissues were processed for histopathology and gene expression profiling. Metastasis was scored 60 days after the removal of transplanted tumors. RESULTS: PyMT tumor fragments and cell suspensions from anterior glands grew faster than posterior tumors in serial passages regardless of the site of implantation. Microarray analysis revealed genetic differences between these tumors. The transplantation was reproducible using anterior tumors from multiple GEM, and tumor growth rate correlated with the number of transplanted cells. Similar morphologic appearances were observed in original and transplanted tumors. Metastasis developed in >90% of mice transplanted with PyMT, 40% with BRCA1/p53(+/-) and wnt1/p53(+/-), and 15% with Her2/neu tumors. Expansion of PyMT and wnt1 tumors by serial transplantation for two passages did not lead to significant changes in gene expression. PyMT-transplanted tumors and anterior tumors of transgenic mice showed similar sensitivities to cyclophosphamide and paclitaxel. CONCLUSIONS: Transplantation of GEM tumors can provide a large cohort of mice bearing mammary tumors at the same stage of tumor development and with defined frequency of metastasis in a well-characterized molecular and genetic background.

VIP receptor antagonists inhibit mammary carcinogenesis in C3(1)SV40T antigen mice.

The effects of a vasoactive intestinal peptide (VIP) receptor antagonist on mammary carcinogenesis were investigated using the C3(1)SV40T antigen (ag) mice. Ten microg/day VIPhybrid (VIPhyb) administered daily subcutaneously increased significantly the survival of C3(1)SV40Tag mice. At 5.2 months, VIPhyb significantly reduced the mammary tumor burden in C3(1)SV40Tag mice relative to control animals. 125I-VIP bound with high affinity to mouse mammary tumor homogenate. Because (Lys15, Arg16, Leu27)VIP1-7GRF8-27 (VPAC1 selective) but not Ro25-1553 (VPAC2 selective) inhibited specific 125I-VIP binding to mammary tumor membranes with high affinity, VPAC1 receptors predominate. By RT-PCR, VPAC1 receptor mRNA was detected in mammary tumors. By Western blot, a major 60 Kdalton band was detected in mammary tumor extracts using VPAC1 receptor antisera. By immunocytochemistry, VPAC1-R immunostaining was detected in the cytosol and plasma membrane but not the nucleus of fixed mammary tumor tissue. Using laser capture microdissected tumor cells and surface enhanced laser desorption/ionization (SELDI) techniques on mammary tumor cells, the proteomic profile was altered in mice treated with VIPhyb. Because VPAC1 receptor antagonists increase the survival and reduce the tumor burden in C3(1)SV40Tag mice, they may function as chemopreventive agents in mammary cancer.

Prostate development and carcinogenesis in prolactin receptor knockout mice.

Hyperprolactinemia results in prostatic hypertrophy and hyperplasia, but it is not known whether prolactin plays an essential role in these processes in the prostate. To address this question, we investigated prostate development, gene expression, and simian virus 40 (SV40)T-induced prostate carcinogenesis in prolactin receptor knockout mice. These animals showed a small increase in dorsolateral and ventral prostate weight but no change in the weight of the anterior prostate. The dorsal but not ventral or lateral lobes showed a 12% loss of epithelial cells; all other morphological parameters were normal. The area of SV40T-induced prostate intraepithelial neoplasia was reduced by 28% in the ventral lobe but not the dorsal lobe, and no tumors were seen in 20 prolactin receptor knockout animals, compared with 1 of 11 detected in wild-type and 4 of 21 found in heterozygous animals. Oligonucleotide microarrays were used to identify essential transcriptional roles of prolactin and revealed a small set of genes with decreased expression involved in sperm/oocyte interaction and copulatory plug formation. Infertility or reduced fertility was apparent in these animals. These findings establish essential though subtle roles for prolactin in the regulation of prostate morphology, gene expression, SV40T-induced neoplasia, and reproductive function.

Suppression of mammary tumorigenesis in transgenic mice by the RXR-selective retinoid, LGD1069.

Retinoids have been used in the clinic for the prevention and treatment of human cancers. They regulate several cellular processes including growth, differentiation, and apoptosis. Previously, we demonstrated that a pan-agonist retinoid 9-cis retinoic acid was able to suppress mammary tumorigenesis in the C3(1)-SV40 T-antigen (Tag) transgenic mouse model. However, significant toxicity was seen with this naturally occurring retinoid. We hypothesized that the cancer preventive effects of retinoids could be dissected from the toxic effects by using receptor-selective retinoids. In this study, we used TTNPB, an retinoic acid receptor-selective retinoid, and LGD1069, an retinoid X receptor-selective retinoid, to preferentially activate retinoic acid receptors and retinoid X receptors. In vitro, both compounds were able to inhibit the growth of T47D breast cancer cells. We then determined whether these retinoids prevented mammary tumorigenesis. C3(1)-SV40 Tag mice were treated daily by gastric gavage with vehicle, two different doses of TTNPB (0.3 or 3.0 microg/kg), or two different doses of LGD1069 (10 or 100 mg/kg). Mice were treated from approximately 6-8 weeks to 7-8 months of age. Tumor size and number were measured twice each week, and toxicities were recorded daily. Our data show that LGD1069 suppresses mammary tumorigenesis in C3(1)-SV40 Tag transgenic mice with no observable toxicity, whereas TTNPB had a modest chemopreventive effect, yet was very toxic. Median time to tumor development was 129 days in vehicle-treated mice versus 156 days in mice treated with 100 mg/kg LGD1069 (P = 0.05). In addition, tumor multiplicity was reduced by approximately 50% in mice treated with LGD1069 (2.9 for vehicle, 2.4 for 10 mg/kg LGD1069, and 1.4 for 100 mg/kg, P < or = 0.03). TTNPB-treated mice showed a delayed median time to tumor development (131 days for vehicle versus 154 days for 3.0 microg/kg TTNPB; P < or = 0.05), but no changes were seen in tumor multiplicity. However, toxicity (skin erythema, hair loss) was seen in all of the mice treated with TTNPB. These data demonstrate that receptor-selective retinoids suppress mammary tumorigenesis in transgenic mice and that preventive effects of retinoids can be separated from their toxicity, demonstrating that receptor-selective retinoids are promising agents for the prevention of breast cancer.