Efficient chemical-disease identification and relationship extraction using Wikipedia to improve recall.

Database (Oxford). 2016;2016:

Authors: Lowe DM, O'Boyle NM, Sayle RA

Abstract
Awareness of the adverse effects of chemicals is important in biomedical research and healthcare. Text mining can allow timely and low-cost extraction of this knowledge from the biomedical literature. We extended our text mining solution, LeadMine, to identify diseases and chemical-induced disease relationships (CIDs). LeadMine is a dictionary/grammar-based entity recognizer and was used to recognize and normalize both chemicals and diseases to Medical Subject Headings (MeSH) IDs. The disease lexicon was obtained from three sources: MeSH, the Disease Ontology and Wikipedia. The Wikipedia dictionary was derived from pages with a disease/symptom box, or those where the page title appeared in the lexicon. Composite entities (e.g. heart and lung disease) were detected and mapped to their composite MeSH IDs. For CIDs, we developed a simple pattern-based system to find relationships within the same sentence. Our system was evaluated in the BioCreative V Chemical-Disease Relation task and achieved very good results for both disease concept ID recognition (F1-score: 86.12%) and CIDs (F1-score: 52.20%) on the test set. As our system was over an order of magnitude faster than other solutions evaluated on the task, we were able to apply the same system to the entirety of MEDLINE allowing us to extract a collection of over 250 000 distinct CIDs.

PMID: 27060160 [PubMed - indexed for MEDLINE]

Novel Therapeutics Identification for Fibrosis in Renal Allograft Using Integrative Informatics Approach.

Sci Rep. 2017 Jan 04;7:39487

Authors: Li L, Greene I, Readhead B, Menon MC, Kidd BA, Uzilov AV, Wei C, Philippe N, Schroppel B, He JC, Chen R, Dudley JT, Murphy B

Abstract
Chronic allograft damage, defined by interstitial fibrosis and tubular atrophy (IF/TA), is a leading cause of allograft failure. Few effective therapeutic options are available to prevent the progression of IF/TA. We applied a meta-analysis approach on IF/TA molecular datasets in Gene Expression Omnibus to identify a robust 85-gene signature, which was used for computational drug repurposing analysis. Among the top ranked compounds predicted to be therapeutic for IF/TA were azathioprine, a drug to prevent acute rejection in renal transplantation, and kaempferol and esculetin, two drugs not previously described to have efficacy for IF/TA. We experimentally validated the anti-fibrosis effects of kaempferol and esculetin using renal tubular cells in vitro and in vivo in a mouse Unilateral Ureteric Obstruction (UUO) model. Kaempferol significantly attenuated TGF-β1-mediated profibrotic pathways in vitro and in vivo, while esculetin significantly inhibited Wnt/β-catenin pathway in vitro and in vivo. Histology confirmed significantly abrogated fibrosis by kaempferol and esculetin in vivo. We developed an integrative computational framework to identify kaempferol and esculetin as putatively novel therapies for IF/TA and provided experimental evidence for their therapeutic activities in vitro and in vivo using preclinical models. The findings suggest that both drugs might serve as therapeutic options for IF/TA.

PMID: 28051114 [PubMed - in process]

Sjögren-Larsson syndrome: a rare disease of the skin and central nervous system.

BMJ Case Rep. 2016 Apr 19;2016:10.1136/bcr-2016-215110

Authors: Roy U, Das U, Pandit A, Debnath A

Abstract
Sjögren-Larsson syndrome is a recessively inherited disease caused by a deficiency of fatty aldehyde dehydrogenase with presenting features of congenital ichthyosis, spastic diplegia or tetraplegia, and mental retardation. The basic pathogenic mechanism is deficiency of fatty aldehyde dehydrogenase, which may lead to an accumulation of long-chain fatty alcohols hampering cell membrane integrity, which further disrupts the barrier function of skin and white matter of the brain. MRI of the brain shows diffuse symmetrical white matter hyperintensities on T2-weighted sequences. Although there is no definitive cure for Sjögren-Larsson syndrome, most patients survive until adulthood and management involves therapies directed towards controlling specific problems. We present a case of Sjögren-Larsson syndrome with classical clinical and MRI features, including a few distinctly atypical characteristics in various attributes.

PMID: 27095813 [PubMed - indexed for MEDLINE]

A rare complication resulting in a rare disease: radiation-induced male breast cancer.

BMJ Case Rep. 2016 Apr 15;2016:10.1136/bcr-2015-211874

Authors: Alazhri J, Saclarides C, Avisar E

Abstract
The increase in survival after childhood radiation therapy for some blood malignancies has led to an increase in the diagnosis of radiation-induced secondary solid malignancies (SSM). We report a young man presenting with invasive breast cancer 19 years after receiving radiation therapy and bone marrow transplant for acute lymphocytic leukaemia in childhood. This latency period is longer than previously reported. Therefore, survivors of radiation-treated primary cancer should be closely monitored for SSM, including breast cancer, for the rest of their lives.

PMID: 27084898 [PubMed - indexed for MEDLINE]

Overlapping 16p13.11 deletion and gain of copies variations associated with childhood onset psychosis include genes with mechanistic implications for autism associated pathways: Two case reports.

Am J Med Genet A. 2016 May;170A(5):1165-73

Authors: Brownstein CA, Kleiman RJ, Engle EC, Towne MC, D'Angelo EJ, Yu TW, Beggs AH, Picker J, Fogler JM, Carroll D, Schmitt RC, Wolff RR, Shen Y, Lip V, Bilguvar K, Kim A, Tembulkar S, O'Donnell K, Gonzalez-Heydrich J

Abstract
Copy number variability at 16p13.11 has been associated with intellectual disability, autism, schizophrenia, epilepsy, and attention-deficit hyperactivity disorder. Adolescent/adult- onset psychosis has been reported in a subset of these cases. Here, we report on two children with CNVs in 16p13.11 that developed psychosis before the age of 7. The genotype and neuropsychiatric abnormalities of these patients highlight several overlapping genes that have possible mechanistic relevance to pathways previously implicated in Autism Spectrum Disorders, including the mTOR signaling and the ubiquitin-proteasome cascades. A careful screening of the 16p13.11 region is warranted in patients with childhood onset psychosis.

PMID: 26887912 [PubMed - indexed for MEDLINE]

[Financing rare or orphan diseases].

Rev Peru Med Exp Salud Publica. 2014 Oct-Dec;31(4):808-9

Authors: Oyola-García A, Lituma-Aguirre D, Honorio-Morales H, Comisión Sectorial Enfermedades Raras o Huérfanas

PMID: 25597742 [PubMed - indexed for MEDLINE]

Integration and Querying of Genomic and Proteomic Semantic Annotations for Biomedical Knowledge Extraction.

IEEE/ACM Trans Comput Biol Bioinform. 2016 Mar-Apr;13(2):209-19

Authors: Masseroli M, Canakoglu A, Ceri S

Abstract
Understanding complex biological phenomena involves answering complex biomedical questions on multiple biomolecular information simultaneously, which are expressed through multiple genomic and proteomic semantic annotations scattered in many distributed and heterogeneous data sources; such heterogeneity and dispersion hamper the biologists' ability of asking global queries and performing global evaluations. To overcome this problem, we developed a software architecture to create and maintain a Genomic and Proteomic Knowledge Base (GPKB), which integrates several of the most relevant sources of such dispersed information (including Entrez Gene, UniProt, IntAct, Expasy Enzyme, GO, GOA, BioCyc, KEGG, Reactome, and OMIM). Our solution is general, as it uses a flexible, modular, and multilevel global data schema based on abstraction and generalization of integrated data features, and a set of automatic procedures for easing data integration and maintenance, also when the integrated data sources evolve in data content, structure, and number. These procedures also assure consistency, quality, and provenance tracking of all integrated data, and perform the semantic closure of the hierarchical relationships of the integrated biomedical ontologies. At http://www.bioinformatics.deib.polimi.it/GPKB/, a Web interface allows graphical easy composition of queries, although complex, on the knowledge base, supporting also semantic query expansion and comprehensive explorative search of the integrated data to better sustain biomedical knowledge extraction.

PMID: 27045824 [PubMed - indexed for MEDLINE]

A knowledgebase of the human Alu repetitive elements.

J Biomed Inform. 2016 Apr;60:77-83

Authors: Mallona I, Jordà M, Peinado MA

Abstract
Alu elements are the most abundant retrotransposons in the human genome with more than one million copies. Alu repeats have been reported to participate in multiple processes related with genome regulation and compartmentalization. Moreover, they have been involved in the facilitation of pathological mutations in many diseases, including cancer. The contribution of Alus and other repeats in genomic regulation is often overlooked because their study poses technical and analytical challenges hardly attainable with conventional strategies. Here we propose the integration of ontology-based semantic methods to query a knowledgebase for the human Alus. The knowledgebase for the human Alus leverages Sequence (SO) and Gene Ontologies (GO) and is devoted to address functional and genetic information in the genomic context of the Alus. For each Alu element, the closest gene and transcript are stored, as well their functional annotation according to GO, the state of the chromatin and the transcription factors binding sites inside the Alu. The model uses Web Ontology Language (OWL) and Semantic Web Rule Language (SWRL). As a case of use and to illustrate the utility of the tool, we have evaluated the epigenetic states of Alu repeats associated with gene promoters according to their transcriptional activity. The ontology is easily extendable, offering a scaffold for the inclusion of new experimental data. The RDF/XML formalization is freely available at http://aluontology.sourceforge.net/.

PMID: 26827622 [PubMed - indexed for MEDLINE]

Psychological and behavioural impact of returning personal results from whole-genome sequencing: the HealthSeq project.

Eur J Hum Genet. 2017 Jan 04;:

Authors: Sanderson SC, Linderman MD, Suckiel SA, Zinberg R, Wasserstein M, Kasarskis A, Diaz GA, Schadt EE

Abstract
Providing ostensibly healthy individuals with personal results from whole-genome sequencing could lead to improved health and well-being via enhanced disease risk prediction, prevention, and diagnosis, but also poses practical and ethical challenges. Understanding how individuals react psychologically and behaviourally will be key in assessing the potential utility of personal whole-genome sequencing. We conducted an exploratory longitudinal cohort study in which quantitative surveys and in-depth qualitative interviews were conducted before and after personal results were returned to individuals who underwent whole-genome sequencing. The participants were offered a range of interpreted results, including Alzheimer's disease, type 2 diabetes, pharmacogenomics, rare disease-associated variants, and ancestry. They were also offered their raw data. Of the 35 participants at baseline, 29 (82.9%) completed the 6-month follow-up. In the quantitative surveys, test-related distress was low, although it was higher at 1-week than 6-month follow-up (Z=2.68, P=0.007). In the 6-month qualitative interviews, most participants felt happy or relieved about their results. A few were concerned, particularly about rare disease-associated variants and Alzheimer's disease results. Two of the 29 participants had sought clinical follow-up as a direct or indirect consequence of rare disease-associated variants results. Several had mentioned their results to their doctors. Some participants felt having their raw data might be medically useful to them in the future. The majority reported positive reactions to having their genomes sequenced, but there were notable exceptions to this. The impact and value of returning personal results from whole-genome sequencing when implemented on a larger scale remains to be seen.European Journal of Human Genetics advance online publication, 4 January 2017; doi:10.1038/ejhg.2016.178.

PMID: 28051073 [PubMed - as supplied by publisher]

A Case of Surgical Site Infection Caused by Mycobacterium fortuitum, following Herniorrhaphy.

J Clin Diagn Res. 2016 Nov;10(11):DD01-DD02

Authors: Madhusudhan NS, Malini A, Sangma MM

Abstract
Rapidly Growing Mycobacteria (RGM) are opportunistic pathogens found in the environment. Mycobacterium fortuitum, M. chelonae and M.abscessus are the important human pathogens of this group. They cause wound infections, disseminated cutaneous disease, pulmonary infection in patients with cystic fibrosis or bronchiectasis, bone and joint infections and keratitis. Infections due to these Non-Tuberculous Mycobacteria (NTM) are increasingly reported. Post laparoscopic wound infections, mesh site infections and other surgical site infections due to M. fortuitum and M. chelonae have been reported. Usually wound infections due to atypical mycobacteria have delayed onset and do not respond to conventional antibiotics. Identification of RGM can be done by a set of cumbersome biochemical tests, High Performance Liquid Chromatography (HPLC), molecular methods using DNA probes or by Polymerase Chain Reaction (PCR). We here report a case of post-herniorrhaphy wound infection due to M. fortuitum which was identified by molecular method (HAIN mycobacterial species system). This case report underscores the importance of examining Ziehl-Neelsen (ZN) stain of all exudates with sterile culture on day one for non fastidious bacteria. Timely identification can lead to prompt therapy of patients preventing further complications.

PMID: 28050369 [PubMed]

Alternative Indexes to Estimate the Functional Capacity From the 6-Minute Walk Test in Children and Adolescents With Cystic Fibrosis.

Respir Care. 2017 Jan 03;:

Authors: Okuro RT, de Oliveira Ribeiro MA, Ribeiro JD, Minsky RC, Schivinski CI

Abstract
BACKGROUND: Cystic fibrosis is a multi-systemic disease related to reduced functional capacity. The distance covered in the 6-min walk test (6MWT) has been known to assess functional capacity, but little is known about other indexes that can be derived. We sought to compare the performance during the 6MWT and the estimated indexes of functional capacity from the 6MWT between subjects with cystic fibrosis (CF) and healthy individuals as well as to assess the relationship among these indexes and disease severity, pulmonary function, and nutritional status in CF.
METHODS: This cross-sectional study was carried out at a university referral center for CF. It included a group of 55 non-oxygen-dependent CF subjects (CF group) with no acute pulmonary exacerbations and a group of 185 healthy controls (control group). All subjects were submitted to 6MWT and anthropometrics measurements.
RESULTS: Regarding performance during the 6MWT, the mean values of work, physiological cost index, average velocity, and 6-min walk distance (6MWD) were significantly lower in the CF group than in the control group (work: 21,690.58 ± 10,427.77 vs 26,057.51 ± 11,228.49 m × kg [P = .007]; physiological cost index: 0.31 ± 0.19 vs 0.37 ± 0.17; average velocity: 94.71 ± 12.89 vs 104.55 ± 9.13 m/min [P < .001]; and 6MWD: 568.02 ± 76.31 m versus 627.54 ± 54.81 m [P < .001]). Subjects with less severe CF had higher 6MWD, work, and average velocity during the 6MWT, compared with subjects with more severe CF (P = .008, P = .012, and P = .007, respectively). There was a correlation between 6MWD, work, average velocity, and disease severity and pulmonary function.
CONCLUSIONS: Considering the importance of standard measure (6MWD) the in 6MWT, alternative indexes can be useful as complementary outcomes and to provide a better understanding of limiting factors of exercise response in children and adolescents with CF.

PMID: 28049743 [PubMed - as supplied by publisher]

Cystic Fibrosis Mutation Detection with SPR Biosensor in Real Samples via Multiple Surfaces Binding Method.

Comb Chem High Throughput Screen. 2017 Jan 03;

Authors: Kayran YU, Ozkan-Ariksoysal D, Topkaya SN, Kaymaz BT, Can BK

Abstract
Surface Plasmon Resonance (SPR) based biosensor system was developed for the detection of Delta F508 (DF508del) Cystic Fibrosis (CF) mutation in both synthetic and real samples. In order to provide an effective hybridization between probe and the Polymerase Chain Reaction (PCR) amplicons (target), streptavidin was bound to the surface and biotin-tag probe was sent to the streptavidin-coated surface. For the target preparation, blood samples were collected from the patients who suffer from CF. Following the DNA isolation; samples were amplified with PCR with biotin-tag. Before sending the biotin-tag PCR amplicons onto the modified surface, amplicons were also interacted with the helper oligonucleotides to prevent re-annealing of the denatured DNA strands. This kind of 'multiple surface binding' method helps increasing of the sensitivity of the detection. The limit of detection(S/N= 3) was calculated as 12.24 pico-mole/ml for PCR-like synthetic long target sequence and 13x105 molecules for real samples in less than half hour. Using the both biotin-tag probe and the helper oligonucleotides together, hybridization was obtained much more efficiently than traditional denaturation protocols for real samples and biotin-free hybridization detection. To the best of our knowledge, the procedure described in this study is one of the simplest, rapid and sensitive methods for CF mutation detection with SPR based biosensor system in real samples.

PMID: 28049404 [PubMed - as supplied by publisher]

Genetic discrimination lawsuit raises broader concerns about testing, privacy: Case involves middle school student impacted by results of genetic screening test as newborn.

Am J Med Genet A. 2016 May;170A(5):1111-2

Authors:

PMID: 27075500 [PubMed - indexed for MEDLINE]

Inhibition of airway surface fluid absorption by cholinergic stimulation.

Sci Rep. 2016 Feb 05;6:20735

Authors: Joo NS, Krouse ME, Choi JY, Cho HJ, Wine JJ

Abstract
In upper airways airway surface liquid (ASL) depth and clearance rates are both increased by fluid secretion. Secretion is opposed by fluid absorption, mainly via the epithelial sodium channel, ENaC. In static systems, increased fluid depth activates ENaC and decreased depth inhibits it, suggesting that secretion indirectly activates ENaC to reduce ASL depth. We propose an alternate mechanism in which cholinergic input, which causes copious airway gland secretion, also inhibits ENaC-mediated absorption. The conjoint action accelerates clearance, and the increased transport of mucus out of the airways restores ASL depth while cleansing the airways. We were intrigued by early reports of cholinergic inhibition of absorption by airways in some species. To reinvestigate this phenomenon, we studied inward short-circuit currents (Isc) in tracheal mucosa from human, sheep, pig, ferret, and rabbit and in two types of cultured cells. Basal Isc was inhibited 20-70% by the ENaC inhibitor, benzamil. Long-lasting inhibition of ENaC-dependent Isc was also produced by basolateral carbachol in all preparations except rabbit and the H441 cell line. Atropine inhibition produced a slow recovery or prevented inhibition if added before carbachol. The mechanism for inhibition was not determined and is most likely multi-factorial. However, its physiological significance is expected to be increased mucus clearance rates in cholinergically stimulated airways.

PMID: 26846701 [PubMed - indexed for MEDLINE]

Modeling Dynamics and Function of Bone Marrow Cells in Mouse Liver Regeneration.

Cell Rep. 2017 Jan 03;18(1):107-121

Authors: Pedone E, Olteanu VA, Marucci L, Muñoz-Martin MI, Youssef SA, de Bruin A, Cosma MP

Abstract
In rodents and humans, the liver can efficiently restore its mass after hepatectomy. This is largely attributed to the proliferation and cell cycle re-entry of hepatocytes. On the other hand, bone marrow cells (BMCs) migrate into the liver after resection. Here, we find that a block of BMC recruitment into the liver severely impairs its regeneration after the surgery. Mobilized hematopoietic stem and progenitor cells (HSPCs) in the resected liver can fuse with hepatocytes, and the hybrids proliferate earlier than the hepatocytes. Genetic ablation of the hybrids severely impairs hepatocyte proliferation and liver mass regeneration. Mathematical modeling reveals a key role of bone marrow (BM)-derived hybrids to drive proliferation in the regeneration process, and predicts regeneration efficiency in experimentally non-testable conditions. In conclusion, BM-derived hybrids are essential to trigger efficient liver regeneration after hepatectomy.

PMID: 28052241 [PubMed - in process]

Seeing the Forest for the Trees.

Circ Res. 2016 Nov 11;119(11):1170-1172

Authors: O'Rourke B, Liu T, Foster DB

PMID: 28051783 [PubMed - in process]

Widespread Historical Contingency in Influenza Viruses.

Genetics. 2017 Jan;205(1):409-420

Authors: Nshogozabahizi JC, Dench J, Aris-Brosou S

Abstract
In systems biology and genomics, epistasis characterizes the impact that a substitution at a particular location in a genome can have on a substitution at another location. This phenomenon is often implicated in the evolution of drug resistance or to explain why particular "disease-causing" mutations do not have the same outcome in all individuals. Hence, uncovering these mutations and their locations in a genome is a central question in biology. However, epistasis is notoriously difficult to uncover, especially in fast-evolving organisms. Here, we present a novel statistical approach that replies on a model developed in ecology and that we adapt to analyze genetic data in fast-evolving systems such as the influenza A virus. We validate the approach using a two-pronged strategy: extensive simulations demonstrate a low-to-moderate sensitivity with excellent specificity and precision, while analyses of experimentally validated data recover known interactions, including in a eukaryotic system. We further evaluate the ability of our approach to detect correlated evolution during antigenic shifts or at the emergence of drug resistance. We show that in all cases, correlated evolution is prevalent in influenza A viruses, involving many pairs of sites linked together in chains; a hallmark of historical contingency. Strikingly, interacting sites are separated by large physical distances, which entails either long-range conformational changes or functional tradeoffs, for which we find support with the emergence of drug resistance. Our work paves a new way for the unbiased detection of epistasis in a wide range of organisms by performing whole-genome scans.

PMID: 28049709 [PubMed - in process]

Verification of a Maternal-Fetal Physiologically Based Pharmacokinetic Model for Passive Placental Permeability Drugs.

Drug Metab Dispos. 2017 Jan 03;:

Authors: Zhang Z, Unadkat JD

Abstract
Fetal exposure to drugs cannot be readily estimated from single time point cord blood sampling at the time of delivery. Therefore, we developed a physiologically-based pharmacokinetic (PBPK) model to estimate fetal drug exposure throughout pregnancy. Here we report verification of this novel maternal-fetal physiologically-based pharmacokinetic (m-f-PBPK) model for drugs that passively diffuse across the placenta and are not metabolized/transported there. Our recently built m-f-PBPK model was populated with gestational age-dependent changes in maternal drug disposition and maternal-fetal physiology. Using midazolam as an in vivo calibrator, the transplacental passive diffusion clearance of theophylline and zidovudine was first estimated. Then, for verification, the predicted maternal plasma (MP) and umbilical venous (UV) plasma drug concentrations by our m-f-PBPK were compared against those observed at term. Overall, our m-f-PBPK model well predicted the maternal and fetal exposure to the two verification drugs, theophylline and zidovudine, at term, across a range of dosing regimens, with nearly all observed MP and UV plasma drug concentrations falling within the 90% prediction interval [i.e.5th -95th percentile range of a virtual pregnant population (n=100)]. Prediction precision and bias of theophylline MP and UV were 14.5% and 12.4%, and 9.4% and 7.5%, respectively. Further, for zidovudine, after the exclusion of one unexpectedly low MP concentration, prediction precision and bias for MP and UV were 50.3 % and 30.2, and 28.3% and 15.0%, respectively. This m-f-PBPK should be useful to predict fetal exposure to drugs, throughout pregnancy, for drugs that passively diffuse across the placenta.

PMID: 28049636 [PubMed - as supplied by publisher]

Natural language processing to ascertain two key variables from operative reports in ophthalmology.

Pharmacoepidemiol Drug Saf. 2017 Jan 03;:

Authors: Liu L, Shorstein NH, Amsden LB, Herrinton LJ

Abstract
PURPOSE: Antibiotic prophylaxis is critical to ophthalmology and other surgical specialties. We performed natural language processing (NLP) of 743 838 operative notes recorded for 315 246 surgeries to ascertain two variables needed to study the comparative effectiveness of antibiotic prophylaxis in cataract surgery. The first key variable was an exposure variable, intracameral antibiotic injection. The second was an intraoperative complication, posterior capsular rupture (PCR), which functioned as a potential confounder. To help other researchers use NLP in their settings, we describe our NLP protocol and lessons learned.
METHODS: For each of the two variables, we used SAS Text Miner and other SAS text-processing modules with a training set of 10 000 (1.3%) operative notes to develop a lexicon. The lexica identified misspellings, abbreviations, and negations, and linked words into concepts (e.g. "antibiotic" linked with "injection"). We confirmed the NLP tools by iteratively obtaining random samples of 2000 (0.3%) notes, with replacement.
RESULTS: The NLP tools identified approximately 60 000 intracameral antibiotic injections and 3500 cases of PCR. The positive and negative predictive values for intracameral antibiotic injection exceeded 99%. For the intraoperative complication, they exceeded 94%.
CONCLUSION: NLP was a valid and feasible method for obtaining critical variables needed for a research study of surgical safety. These NLP tools were intended for use in the study sample. Use with external datasets or future datasets in our own setting would require further testing. Copyright © 2017 John Wiley & Sons, Ltd.

PMID: 28052483 [PubMed - as supplied by publisher]

An Evolving Ecosystem for Natural Language Processing in Department of Veterans Affairs.

J Med Syst. 2017 Feb;41(2):32

Authors: Garvin JH, Kalsy M, Brandt C, Luther SL, Divita G, Coronado G, Redd D, Christensen C, Hill B, Kelly N, Treitler QZ

Abstract
In an ideal clinical Natural Language Processing (NLP) ecosystem, researchers and developers would be able to collaborate with others, undertake validation of NLP systems, components, and related resources, and disseminate them. We captured requirements and formative evaluation data from the Veterans Affairs (VA) Clinical NLP Ecosystem stakeholders using semi-structured interviews and meeting discussions. We developed a coding rubric to code interviews. We assessed inter-coder reliability using percent agreement and the kappa statistic. We undertook 15 interviews and held two workshop discussions. The main areas of requirements related to; design and functionality, resources, and information. Stakeholders also confirmed the vision of the second generation of the Ecosystem and recommendations included; adding mechanisms to better understand terms, measuring collaboration to demonstrate value, and datasets/tools to navigate spelling errors with consumer language, among others. Stakeholders also recommended capability to: communicate with developers working on the next version of the VA electronic health record (VistA Evolution), provide a mechanism to automatically monitor download of tools and to automatically provide a summary of the downloads to Ecosystem contributors and funders. After three rounds of coding and discussion, we determined the percent agreement of two coders to be 97.2% and the kappa to be 0.7851. The vision of the VA Clinical NLP Ecosystem met stakeholder needs. Interviews and discussion provided key requirements that inform the design of the VA Clinical NLP Ecosystem.

PMID: 28050745 [PubMed - in process]