Using a chat-based informed consent tool in large-scale genomic research

OBJECTIVE

We implemented a chatbot consent tool to shift the time burden from study staff in support of a national genomics research study.

MATERIALS AND METHODS

We created an Institutional Review Board-approved script for automated chat-based consent. We compared data from prospective participants who used the tool or had traditional consent conversations with study staff.

RESULTS

Chat-based consent, completed on a user's schedule, was shorter than the traditional conversation. This did not lead to a significant change in affirmative consents. Within affirmative consents and declines, more prospective participants completed the chat-based process. A quiz to assess chat-based consent user understanding had a high pass rate with no reported negative experiences.

CONCLUSION

Our report shows that a structured script can convey important information while realizing the benefits of automation and burden shifting. Analysis suggests that it may be advantageous to use chatbots to scale this rate-limiting step in large research projects.

Increased diagnostic yield from negative whole genome-slice panels using automated reanalysis

We evaluated the diagnostic yield using genome-slice panel reanalysis in the clinical setting using an automated phenotype/gene ranking system. We analyzed whole genome sequencing (WGS) data produced from clinically ordered panels built as bioinformatic slices for 16 clinically diverse, undiagnosed cases referred to the Pediatric Mendelian Genomics Research Center, an NHGRI-funded GREGoR Consortium site. Genome-wide reanalysis was performed using Moon™, a machine-learning-based tool for variant prioritization. In five out of 16 cases, we discovered a potentially clinically significant variant. In four of these cases, the variant was found in a gene not included in the original panel due to phenotypic expansion of a disorder or incomplete initial phenotyping of the patient. In the fifth case, the gene containing the variant was included in the original panel, but being a complex structural rearrangement with intronic breakpoints outside the clinically analyzed regions, it was not initially identified. Automated genome-wide reanalysis of clinical WGS data generated during targeted panels testing yielded a 25% increase in diagnostic findings and a possibly clinically relevant finding in one additional case, underscoring the added value of analyses versus those routinely performed in the clinical setting.

"Development and Implementation of Novel Chatbot-based Genomic Research Consent"

Objective

To conduct a retrospective analysis comparing traditional human-based consenting to an automated chat-based consenting process.

Materials and Methods

We developed a new chat-based consent using our IRB-approved consent forms. We leveraged a previously developed platform (GiaⓇ, or "Genetic Information Assistant") to deliver the chat content to candidate participants. The content included information about the study, educational information, and a quiz to assess understanding. We analyzed 144 families referred to our study during a 6-month time period. A total of 37 families completed consent using the traditional process, while 35 families completed consent using Gia.

Results

Engagement rates were similar between both consenting methods. The median length of the consent conversation was shorter for Gia users compared to traditional (44 vs. 76 minutes). Additionally, the total time from referral to consent completion was faster with Gia (5 vs. 16 days). Within Gia, understanding was assessed with a 10-question quiz that most participants (96%) passed. Feedback about the chat consent indicated that 86% of participants had a positive experience.

Discussion

Using Gia resulted in time savings for both the participant and study staff. The chatbot enables studies to reach more potential candidates. We identified five key features related to human-centered design for developing a consent chat.

Conclusion

This analysis suggests that it is feasible to use an automated chatbot to scale obtaining informed consent for a genomics research study. We further identify a number of advantages when using a chatbot.

State Variation in TANF Expenditures: Implications for Social Work and Social Policy

The Temporary Assistance for Needy Families (TANF) program is a federal block grant to the states, with a required state contribution. Although often viewed as a cash assistance program with work requirements and services targeted at extremely low-income families with children, only about one-quarter of all state and federal TANF funds are now used for traditional cash aid. Uses of funds vary widely by state, and alternatives range from refundable tax credits to support of state child welfare systems. In this article, the author examines the relationship between state categorical TANF spending and key social, political, and economic characteristics using data from 2015 to 2017 and multilevel linear models. Racial and ethnic demographics of the cash assistance caseload are associated with differences in spending, with states with larger proportions of the caseload composed of people of color devoting a lower percentage of effort to traditional benefits and more to alternative cash transfers. Changes in unemployment rate within states are associated with greater spending on basic assistance and reduced spending on alternative transfers. These findings indicate that, although TANF cash benefits spending may be economically responsive within the program's overall flexible structure, spending patterns raise issues of equity for disadvantaged families.

Racial and Ethnic Disparities in the Lifetime Prevalence of Homelessness in the United States

Homelessness in the United States is often examined using cross-sectional, point-in-time samples. Any experience of homelessness is a risk factor for adverse outcomes, so it is also useful to understand the incidence of homelessness over longer periods. We estimate the lifetime prevalence of homelessness among members of the Baby Boom cohort (n = 6,545) using the 2012 and 2014 waves of the Health and Retirement Study (HRS), a nationally representative survey of older Americans. Our analysis indicates that 6.2 % of respondents had a period of homelessness at some point in their lives. We also identify dramatic disparities in lifetime incidence of homelessness by racial and ethnic subgroups. Rates of homelessness were higher for non-Hispanic blacks (16.8 %) or Hispanics of any race (8.1 %) than for non-Hispanic whites (4.8 %; all differences significant with p < .05). The black-white gap, but not the Hispanic-white gap, remained significant after adjustment for covariates such as education, veteran status, and geographic region.

Creating a scalable clinical pharmacogenomics service with automated interpretation and medical record result integration - experience from a pediatric tertiary care facility

OBJECTIVE

This paper outlines the implementation of a comprehensive clinical pharmacogenomics (PGx) service within a pediatric teaching hospital and the integration of clinical decision support in the electronic health record (EHR).

MATERIALS AND METHODS

An approach to clinical decision support for medication ordering and dispensing driven by documented PGx variant status in an EHR is described. A web-based platform was created to automatically generate a clinical report from either raw assay results or specified diplotypes, able to parse and combine haplotypes into an interpretation for each individual and compared to the reference lab call for accuracy.

RESULTS

Clinical decision support rules built within an EHR provided guidance to providers for 31 patients (100%) who had actionable PGx variants and were written for interacting medications. A breakdown of the PGx alerts by practitioner service, and alert response for the initial cohort of patients tested is described. In 90% (355/394) of the cases, thiopurine methyltranferase genotyping was ordered pre-emptively.

DISCUSSION

This paper outlines one approach to implementing a clinical PGx service in a pediatric teaching hospital that cares for a heterogeneous patient population. There is a focus on incorporation of PGx clinical decision support rules and a program to standardize report text within the electronic health record with subsequent exploration of clinician behavior in response to the alerts.

CONCLUSION

The incorporation of PGx data at the time of prescribing and dispensing, if done correctly, has the potential to impact the incidence of adverse drug events, a significant cause of morbidity and mortality.

A Python package for parsing, validating, mapping and formatting sequence variants using HGVS nomenclature

Biological sequence variants are commonly represented in scientific literature, clinical reports and databases of variation using the mutation nomenclature guidelines endorsed by the Human Genome Variation Society (HGVS). Despite the widespread use of the standard, no freely available and comprehensive programming libraries are available. Here we report an open-source and easy-to-use Python library that facilitates the parsing, manipulation, formatting and validation of variants according to the HGVS specification. The current implementation focuses on the subset of the HGVS recommendations that precisely describe sequence-level variation relevant to the application of high-throughput sequencing to clinical diagnostics.

AVAILABILITY AND IMPLEMENTATION

The package is released under the Apache 2.0 open-source license. Source code, documentation and issue tracking are available at http://bitbucket.org/hgvs/hgvs/. Python packages are available at PyPI (https://pypi.python.org/pypi/hgvs).

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

The potential of accelerating early detection of autism through content analysis of YouTube videos

Autism is on the rise, with 1 in 88 children receiving a diagnosis in the United States, yet the process for diagnosis remains cumbersome and time consuming. Research has shown that home videos of children can help increase the accuracy of diagnosis. However the use of videos in the diagnostic process is uncommon. In the present study, we assessed the feasibility of applying a gold-standard diagnostic instrument to brief and unstructured home videos and tested whether video analysis can enable more rapid detection of the core features of autism outside of clinical environments. We collected 100 public videos from YouTube of children ages 1–15 with either a self-reported diagnosis of an ASD (N = 45) or not (N = 55). Four non-clinical raters independently scored all videos using one of the most widely adopted tools for behavioral diagnosis of autism, the Autism Diagnostic Observation Schedule-Generic (ADOS). The classification accuracy was 96.8%, with 94.1% sensitivity and 100% specificity, the inter-rater correlation for the behavioral domains on the ADOS was 0.88, and the diagnoses matched a trained clinician in all but 3 of 22 randomly selected video cases. Despite the diversity of videos and non-clinical raters, our results indicate that it is possible to achieve high classification accuracy, sensitivity, and specificity as well as clinically acceptable inter-rater reliability with nonclinical personnel. Our results also demonstrate the potential for video-based detection of autism in short, unstructured home videos and further suggests that at least a percentage of the effort associated with detection and monitoring of autism may be mobilized and moved outside of traditional clinical environments.

Early detection of poor adherers to statins: applying individualized surveillance to pay for performance

BACKGROUND

Medication nonadherence costs $300 billion annually in the US. Medicare Advantage plans have a financial incentive to increase medication adherence among members because the Centers for Medicare and Medicaid Services (CMS) now awards substantive bonus payments to such plans, based in part on population adherence to chronic medications. We sought to build an individualized surveillance model that detects early which beneficiaries will fall below the CMS adherence threshold.

METHODS

This was a retrospective study of over 210,000 beneficiaries initiating statins, in a database of private insurance claims, from 2008-2011. A logistic regression model was constructed to use statin adherence from initiation to day 90 to predict beneficiaries who would not meet the CMS measure of proportion of days covered 0.8 or above, from day 91 to 365. The model controlled for 15 additional characteristics. In a sensitivity analysis, we varied the number of days of adherence data used for prediction.

RESULTS

Lower adherence in the first 90 days was the strongest predictor of one-year nonadherence, with an odds ratio of 25.0 (95% confidence interval 23.7-26.5) for poor adherence at one year. The model had an area under the receiver operating characteristic curve of 0.80. Sensitivity analysis revealed that predictions of comparable accuracy could be made only 40 days after statin initiation. When members with 30-day supplies for their first statin fill had predictions made at 40 days, and members with 90-day supplies for their first fill had predictions made at 100 days, poor adherence could be predicted with 86% positive predictive value.

CONCLUSIONS

To preserve their Medicare Star ratings, plan managers should identify or develop effective programs to improve adherence. An individualized surveillance approach can be used to target members who would most benefit, recognizing the tradeoff between improved model performance over time and the advantage of earlier detection.

Development of a scalable pharmacogenomic clinical decision support service

Advances in sequencing technology are making genomic data more accessible within the healthcare environment. Published pharmacogenetic guidelines attempt to provide a clinical context for specific genomic variants; however, the actual implementation to convert genomic data into a clinical report integrated within an electronic medical record system is a major challenge for any hospital. We created a two-part solution that integrates with the medical record system and converts genetic variant results into an interpreted clinical report based on published guidelines. We successfully developed a scalable infrastructure to support TPMT genetic testing and are currently testing approximately two individuals per week in our production version. We plan to release an online variant to clinical interpretation reporting system in order to facilitate translation of pharmacogenetic information into clinical practice.

A systems approach to designing effective clinical trials using simulations

BACKGROUND

Pharmacogenetics in warfarin clinical trials have failed to show a significant benefit in comparison with standard clinical therapy. This study demonstrates a computational framework to systematically evaluate preclinical trial design of target population, pharmacogenetic algorithms, and dosing protocols to optimize primary outcomes.

METHODS AND RESULTS

We programmatically created an end-to-end framework that systematically evaluates warfarin clinical trial designs. The framework includes options to create a patient population, multiple dosing strategies including genetic-based and nongenetic clinical-based, multiple-dose adjustment protocols, pharmacokinetic/pharmacodynamics modeling and international normalization ratio prediction, and various types of outcome measures. We validated the framework by conducting 1000 simulations of the applying pharmacogenetic algorithms to individualize dosing of warfarin (CoumaGen) clinical trial primary end points. The simulation predicted a mean time in therapeutic range of 70.6% and 72.2% (P=0.47) in the standard and pharmacogenetic arms, respectively. Then, we evaluated another dosing protocol under the same original conditions and found a significant difference in the time in therapeutic range between the pharmacogenetic and standard arm (78.8% versus 73.8%; P=0.0065), respectively.

CONCLUSIONS

We demonstrate that this simulation framework is useful in the preclinical assessment phase to study and evaluate design options and provide evidence to optimize the clinical trial for patient efficacy and reduced risk.

Use of machine learning to shorten observation-based screening and diagnosis of autism

The Autism Diagnostic Observation Schedule-Generic (ADOS) is one of the most widely used instruments for behavioral evaluation of autism spectrum disorders. It is composed of four modules, each tailored for a specific group of individuals based on their language and developmental level. On average, a module takes between 30 and 60 min to deliver. We used a series of machine-learning algorithms to study the complete set of scores from Module 1 of the ADOS available at the Autism Genetic Resource Exchange (AGRE) for 612 individuals with a classification of autism and 15 non-spectrum individuals from both AGRE and the Boston Autism Consortium (AC). Our analysis indicated that 8 of the 29 items contained in Module 1 of the ADOS were sufficient to classify autism with 100% accuracy. We further validated the accuracy of this eight-item classifier against complete sets of scores from two independent sources, a collection of 110 individuals with autism from AC and a collection of 336 individuals with autism from the Simons Foundation. In both cases, our classifier performed with nearly 100% sensitivity, correctly classifying all but two of the individuals from these two resources with a diagnosis of autism, and with 94% specificity on a collection of observed and simulated non-spectrum controls. The classifier contained several elements found in the ADOS algorithm, demonstrating high test validity, and also resulted in a quantitative score that measures classification confidence and extremeness of the phenotype. With incidence rates rising, the ability to classify autism effectively and quickly requires careful design of assessment and diagnostic tools. Given the brevity, accuracy and quantitative nature of the classifier, results from this study may prove valuable in the development of mobile tools for preliminary evaluation and clinical prioritization-in particular those focused on assessment of short home videos of children--that speed the pace of initial evaluation and broaden the reach to a significantly larger percentage of the population at risk.

A simulation platform to examine heterogeneity influence on treatment

Although a protocol aims to guide treatment management and optimize overall outcomes, the benefits and harms for each individual vary due to heterogeneity. Some protocols integrate clinical and genetic variation to provide treatment recommendation; it is not clear whether such integration is sufficient. If not, treatment outcomes may be sub-optimal for certain patient sub-populations. Unfortunately, running a clinical trial to examine such outcome responses is cost prohibitive and requires a significant amount of time to conduct the study. We propose a simulation approach to discover this knowledge from electronic medical records; a rapid method to reach this goal. We use the well-known drug warfarin as an example to examine whether patient characteristics, including race and the genes CYP2C9 and VKORC1, have been fully integrated into dosing protocols. The two genes mentioned above have been shown to be important in patient response to warfarin.

Cost-effective cloud computing: a case study using the comparative genomics tool, roundup

Background

Comparative genomics resources, such as ortholog detection tools and repositories are rapidly increasing in scale and complexity. Cloud computing is an emerging technological paradigm that enables researchers to dynamically build a dedicated virtual cluster and may represent a valuable alternative for large computational tools in bioinformatics. In the present manuscript, we optimize the computation of a large-scale comparative genomics resource—Roundup—using cloud computing, describe the proper operating principles required to achieve computational efficiency on the cloud, and detail important procedures for improving cost-effectiveness to ensure maximal computation at minimal costs.

Methods

Utilizing the comparative genomics tool, Roundup, as a case study, we computed orthologs among 902 fully sequenced genomes on Amazon’s Elastic Compute Cloud. For managing the ortholog processes, we designed a strategy to deploy the web service, Elastic MapReduce, and maximize the use of the cloud while simultaneously minimizing costs. Specifically, we created a model to estimate cloud runtime based on the size and complexity of the genomes being compared that determines in advance the optimal order of the jobs to be submitted.

Results

We computed orthologous relationships for 245,323 genome-to-genome comparisons on Amazon’s computing cloud, a computation that required just over 200 hours and cost $8,000 USD, at least 40% less than expected under a strategy in which genome comparisons were submitted to the cloud randomly with respect to runtime. Our cost savings projections were based on a model that not only demonstrates the optimal strategy for deploying RSD to the cloud, but also finds the optimal cluster size to minimize waste and maximize usage. Our cost-reduction model is readily adaptable for other comparative genomics tools and potentially of significant benefit to labs seeking to take advantage of the cloud as an alternative to local computing infrastructure.

Immunoaffinity enrichment and liquid chromatography-selected reaction monitoring mass spectrometry for quantitation of carbonic anhydrase 12 in cultured renal carcinoma cells

Liquid chromatography-selected reaction monitoring (LC-SRM) is a highly specific and sensitive mass spectrometry (MS) technique that is widely being applied to selectively qualify and validate candidate markers within complex biological samples. However, in order for LC-SRM methods to take on these attributes, target-specific optimization of sample processing is required, in order to reduce analyte complexity, prior to LC-SRM. In this study, we have developed a targeted platform consisting of protein immunoaffinity enrichment on magnetic beads and LC-SRM for measuring carbonic anhydrase 12 (CA12) protein in a renal cell carcinoma (RCC) cell line (PRC3), a candidate biomarker for RCC whose expression at the protein level has not been previously reported. Sample processing and LC-SRM assay were optimized for signature peptides selected as surrogate markers of CA12 protein. Using LC-SRM coupled with stable isotope dilution, we achieved limits of quantitation in the low fmol range sufficient for measuring clinically relevant biomarkers with good intra- and interassay accuracy and precision (≤17%). Our results show that using a quantitative immunoaffinity capture approach provides specific, accurate, and robust assays amenable to high-throughput verification of potential biomarkers.

Genotator: a disease-agnostic tool for genetic annotation of disease

BACKGROUND

Disease-specific genetic information has been increasing at rapid rates as a consequence of recent improvements and massive cost reductions in sequencing technologies. Numerous systems designed to capture and organize this mounting sea of genetic data have emerged, but these resources differ dramatically in their disease coverage and genetic depth. With few exceptions, researchers must manually search a variety of sites to assemble a complete set of genetic evidence for a particular disease of interest, a process that is both time-consuming and error-prone.

METHODS

We designed a real-time aggregation tool that provides both comprehensive coverage and reliable gene-to-disease rankings for any disease. Our tool, called Genotator, automatically integrates data from 11 externally accessible clinical genetics resources and uses these data in a straightforward formula to rank genes in order of disease relevance. We tested the accuracy of coverage of Genotator in three separate diseases for which there exist specialty curated databases, Autism Spectrum Disorder, Parkinson's Disease, and Alzheimer Disease. Genotator is freely available at http://genotator.hms.harvard.edu.

RESULTS

Genotator demonstrated that most of the 11 selected databases contain unique information about the genetic composition of disease, with 2514 genes found in only one of the 11 databases. These findings confirm that the integration of these databases provides a more complete picture than would be possible from any one database alone. Genotator successfully identified at least 75% of the top ranked genes for all three of our use cases, including a 90% concordance with the top 40 ranked candidates for Alzheimer Disease.

CONCLUSIONS

As a meta-query engine, Genotator provides high coverage of both historical genetic research as well as recent advances in the genetic understanding of specific diseases. As such, Genotator provides a real-time aggregation of ranked data that remains current with the pace of research in the disease fields. Genotator's algorithm appropriately transforms query terms to match the input requirements of each targeted databases and accurately resolves named synonyms to ensure full coverage of the genetic results with official nomenclature. Genotator generates an excel-style output that is consistent across disease queries and readily importable to other applications.

Cloud computing for comparative genomics

Background

Large comparative genomics studies and tools are becoming increasingly more compute-expensive as the number of available genome sequences continues to rise. The capacity and cost of local computing infrastructures are likely to become prohibitive with the increase, especially as the breadth of questions continues to rise. Alternative computing architectures, in particular cloud computing environments, may help alleviate this increasing pressure and enable fast, large-scale, and cost-effective comparative genomics strategies going forward. To test this, we redesigned a typical comparative genomics algorithm, the reciprocal smallest distance algorithm (RSD), to run within Amazon's Elastic Computing Cloud (EC2). We then employed the RSD-cloud for ortholog calculations across a wide selection of fully sequenced genomes.

Results

We ran more than 300,000 RSD-cloud processes within the EC2. These jobs were farmed simultaneously to 100 high capacity compute nodes using the Amazon Web Service Elastic Map Reduce and included a wide mix of large and small genomes. The total computation time took just under 70 hours and cost a total of $6,302 USD.

Conclusions

The effort to transform existing comparative genomics algorithms from local compute infrastructures is not trivial. However, the speed and flexibility of cloud computing environments provides a substantial boost with manageable cost. The procedure designed to transform the RSD algorithm into a cloud-ready application is readily adaptable to similar comparative genomics problems.

Many Microbe Microarrays Database: uniformly normalized Affymetrix compendia with structured experimental metadata

Many Microbe Microarrays Database (M^3D) is designed to facilitate the analysis and visualization of expression data in compendia compiled from multiple laboratories. M^3D contains over a thousand Affymetrix microarrays for Escherichia coli, Saccharomyces cerevisiae and Shewanella oneidensis. The expression data is uniformly normalized to make the data generated by different laboratories and researchers more comparable. To facilitate computational analyses, M^3D provides raw data (CEL file) and normalized data downloads of each compendium. In addition, web-based construction, visualization and download of custom datasets are provided to facilitate efficient interrogation of the compendium for more focused analyses. The experimental condition metadata in M^3D is human curated with each chemical and growth attribute stored as a structured and computable set of experimental features with consistent naming conventions and units. All versions of the normalized compendia constructed for each species are maintained and accessible in perpetuity to facilitate the future interpretation and comparison of results published on M^3D data. M^3D is accessible at http://m3d.bu.edu/.

Nanoporous alumina capsules for cellular macroencapsulation: transport and biocompatibility

BACKGROUND

Currently, the most widely used treatment for diabetes is the daily subcutaneous injection of recombinant human insulin. Daily injections, however, cannot match the physiological biphasic behavior of normal insulin release, nor can they precisely meet the demands of food intake, exercise, and stress. Cellular encapsulation, or immunoisolation, is a possible solution to this problem. This cell-based therapy allows patients to receive the benefits of more physiological insulin and blood glucose regulation, without the need for immunosuppressants that are associated with organ or cell transplantation.

METHODS

Immunoisolation capsules were fabricated out of aluminum and aluminum oxide using a two-step anodization procedure. The diffusion behavior of glucose, immunoglobulin G (IgG), and insulin were measured. Furthermore, the functionality and viability of encapsulated MIN6 cells were tested. Finally, live cells were stained and imaged using confocal microscopy.

RESULTS

Data indicated that this device is effective in allowing the transport of relevant molecules such as glucose and insulin, while at the same time significantly impeding the transport of IgG, suggesting that it would be efficacious in protecting cell grafts in vivo. Furthermore, encapsulated cells were able to respond dynamically to glucose input signals. Finally, cell staining showed that the viability of encapsulated cells is maintained after 24 h, although the cells appear to be more heavily concentrated at the area that is closest to the membrane.

CONCLUSIONS

This study has shown that nanoporous alumina membranes, with well-controlled pore sizes, can be used for the encapsulation of therapeutic cells and may provide an alternative encapsulation strategy for the treatment of diabetes.

Low molecular weight proteomic information distinguishes metastatic from benign pheochromocytoma

Metastatic lesions occur in up to 36% of patients with pheochromocytoma. Currently there is no way to reliably detect or predict which patients are at risk for metastatic pheochromocytoma. Thus, the discovery of biomarkers that could distinguish patients with benign disease from those with metastatic disease would be of great clinical value. Using surface-enhanced laser desorption ionization protein chips combined with high-resolution mass spectrometry, we tested the hypothesis that pheochromocytoma pathologic states can be reflected as biomarker information within the low molecular weight (LMW) region of the serum proteome. LMW protein profiles were generated from the serum of 67 pheochromocytoma patients from four institutions and analyzed by two different bioinformatics approaches employing pattern recognition algorithms to determine if the LMW component of the circulatory proteome contains potentially useful discriminatory information. Both approaches were able to identify combinations of LMW molecules which could distinguish all metastatic from all benign pheochromocytomas in a separate blinded validation set. In conclusion, for this study set low molecular mass biomarker information correlated with pheochromocytoma pathologic state using blinded validation. If confirmed in larger validation studies, efforts to identify the underlying diagnostic molecules by sequencing would be warranted. In the future, measurement of these biomarkers could be potentially used to improve the ability to identify patients with metastatic disease.

Serum proteomic profiling can discriminate prostate cancer from benign prostates in men with total prostate specific antigen levels between 2.5 and 15.0 ng/ml

PURPOSE

Artificial intelligence based pattern recognition algorithms have been developed and successfully used to analyze complex serum proteomic data streams generated by surface enhanced, laser desorption ionization time-of-flight mass spectroscopy. In the current study we used a high performance, hybrid quadrupole time-of-flight mass spectrometer to generate discriminatory serum proteomic profiles to determine if this technology could be used to determine the need for prostate biopsy in men with elevated prostate specific antigen (PSA).

MATERIALS AND METHODS

Serum samples were collected from 154 men with serum PSA 2.5 to 15.0 ng/ml and/or abnormal digital rectal examination prior to transrectal ultrasound guided biopsy. Serum samples were applied to WCX2 (weak cation exchange protein chip) Protein Arrays (Ciphergen Biosystems, Fremont, California) by a Biomek 2000 robotic liquid handler (Beckman-Coulter, Chaska, Minnesota) and low molecular weight (less than 20 kDa) proteomic patterns were generated with an API QSTAR Pulsar i LC/MS/MS System (Applied Biosystems, Framingham, Massachusetts). High resolution mass spectra were analyzed with a pattern recognition bioinformatics tool, that is Proteome Quest beta version 1.0 (Correlogic Systems, Inc., Bethesda, Maryland), in an attempt to identify and discover key discriminating ion signatures. Serum samples from 63 men (2 or more negative prostate biopsies in 23, 1 negative biopsy in 10 and biopsy detected prostate cancer [CaP] in 30) were used to train the diagnostic algorithm. The remaining 91 samples, including 28 of prostate cancer and 63 of 1 or more negative biopsies, were analyzed in blinded fashion.

RESULTS

The most discriminatory model was found using the WCX2 chip. Testing the remaining 91 men with this model yielded 100% sensitivity and 67% specificity. In other words, if the proteomic pattern had been used to determine the need for prostate biopsy in this cohort of men with PSA between 2.5 and 15.0 ng/ml, 67% (42 of 63) with negative biopsies would have avoided unnecessary biopsy, while no cancers would have been missed.

CONCLUSIONS

Our data demonstrate that high resolution mass spectroscopy can generate serum proteomic patterns that discriminate men with elevated PSA due to benign processes from men with CaP even when PSA is within the diagnostic gray zone. We are currently expanding the testing set to determine the reliability of this new technology to decrease unnecessary prostate biopsies without compromising the detection of curable CaP.

Proteomic analysis of lymph

This report provides the first proteomic analysis of normal ovine lymph. By establishing the fact that lymph is more than an ultrafiltrate of blood plasma, it documents that the lymph proteome contains an array of proteins that differentiates it from plasma. The protein chip technology, surface-enhanced laser desorption/ionization-time of flight-mass spectrometry (SELDI-TOF-MS), two-dimensional gel electrophoresis (2-D PAGE) and MS, were employed to examine the protein expression profiles of ovine lymph. Using a weak cation exchange chip surface to assay lymph and plasma samples by SELDI-TOF-MS showed that the analysis of peak maps from lymph contained three protein peaks that were found only in lymph, while analysis of peak maps from plasma samples showed that five protein peaks were found only in plasma. Lymph and plasma samples showed eight peaks that were common to both. There were also more ions present in plasma than in lymph, which is consistent with the 2-D PAGE analysis. MS analysis of a large number of protein spots from 2-D PAGE gels of lymph produced MS/MS sequences for 18 proteins that were identified by searching against a comprehensive protein sequence database. As in plasma, large protein spots of albumin dominated the protein pattern in lymph. Other major proteins identified in 2-D PAGE gels of lymph included, fibrinogen alpha- and beta-chains, immunoglobulin G (IgG) heavy chain, serotransferrin precursor, lactoferrin, and apolipoprotein A-1. Two proteins that were identified and were differentially expressed in lymph were glial fibrillary astrocyte acidic protein and neutrophil cytosol factor-1. By bringing the technologies of proteomics to bear on the analysis of lymph, it is possible to detect proteins in lymph that are quantitatively and qualitatively differentially expressed from those of plasma.

Novel approaches to visualization and data mining reveals diagnostic information in the low amplitude region of serum mass spectra from ovarian cancer patients

The ability to identify patterns of diagnostic signatures in proteomic data generated by high throughput mass spectrometry (MS) based serum analysis has recently generated much excitement and interest from the scientific community. These data sets can be very large, with high-resolution MS instrumentation producing 1–2 million data points per sample. Approaches to analyze mass spectral data using unsupervised and supervised data mining operations would greatly benefit from tools that effectively allow for data reduction without losing important diagnostic information. In the past, investigators have proposed approaches where data reduction is performed by a priori “peak picking” and alignment/warping/smoothing components using rule-based signal-to-noise measurements. Unfortunately, while this type of system has been employed for gene microarray analysis, it is unclear whether it will be effective in the analysis of mass spectral data, which unlike microarray data, is comprised of continuous measurement operations. Moreover, it is unclear where true signal begins and noise ends. Therefore, we have developed an approach to MS data analysis using new types of data visualization and mining operations in which data reduction is accomplished by culling via the intensity of the peaks themselves instead of by location. Applying this new analysis method on a large study set of high resolution mass spectra from healthy and ovarian cancer patients, shows that all of the diagnostic information is contained within the very lowest amplitude regions of the mass spectra. This region can then be selected and studied to identify the exact location and amplitude of the diagnostic biomarkers.

High-resolution serum proteomic features for ovarian cancer detection

Serum proteomic pattern diagnostics is an emerging paradigm employing low-resolution mass spectrometry (MS) to generate a set of biomarker classifiers. In the present study, we utilized a well-controlled ovarian cancer serum study set to compare the sensitivity and specificity of serum proteomic diagnostic patterns acquired using a high-resolution versus a low-resolution MS platform. In blinded testing sets, the high-resolution mass spectral data contained multiple diagnostic signatures that were superior to the low-resolution spectra in terms of sensitivity and specificity (P<0.00001) throughout the range of modeling conditions. Four mass spectral feature set patterns acquired from data obtained exclusively with the high-resolution mass spectrometer were 100% specific and sensitive in their diagnosis of serum samples as being acquired from either unaffected patients or those suffering from ovarian cancer. Important to the future of proteomic pattern diagnostics is the ability to recognize inferior spectra statistically, so that those resulting from a specific process error are recognized prior to their potentially incorrect (and damaging) diagnosis. To meet this need, we have developed a series of quality-assurance and in-process control procedures to (a) globally evaluate sources of sample variability, (b) identify outlying mass spectra, and (c) develop quality-control release specifications. From these quality-assurance and control (QA/QC) specifications, we identified 32 mass spectra out of the total 248 that showed statistically significant differences from the norm. Hence, 216 of the initial 248 high-resolution mass spectra were determined to be of high quality and were remodeled by pattern-recognition analysis. Again, we obtained four mass spectral feature set patterns that also exhibited 100% sensitivity and specificity in blinded validation tests (68/68 cancer: including 18/18 stage I, and 43/43 healthy). We conclude that (a) the use of high-resolution MS yields superior classification patterns as compared with those obtained with lower resolution instrumentation; (b) although the process error that we discovered did not have a deleterious impact on the present results obtained from proteomic pattern analysis, the major source of spectral variability emanated from mass spectral acquisition, and not bias at the clinical collection site; (c) this variability can be reduced and monitored through the use of QA/QC statistical procedures; (d) multiple and distinct proteomic patterns, comprising low molecular weight biomarkers, detected by high-resolution MS achieve accuracies surpassing individual biomarkers, warranting validation in a large clinical study.

Mass spectrometry-based proteomics and analyses of serum: a primer for the clinical investigator

The vocabulary of proteomics and the swiftly-developing, technological nature of the field constitute substantial barriers to clinical investigators. In recent years, mass spectrometry has emerged as the most promising technique in this field. The purpose of this review is to introduce the field of mass spectrometry-based proteomics to clinical investigators, to explain many of the relevant terms, to introduce the equipment employed in this field, and to outline approaches to asking clinical questions using a proteomic approach. Examples of clinical applications of proteomic techniques are provided from the fields of cancer and vasculitis research, with an emphasis on a pattern recognition approach.

Proteomic technologies to study diseases of the lymphatic vascular system

Now that the human genome has been mapped, a new challenge has emerged: deciphering the various products of individual genes. Consequently, new proteomic technologies are being developed to monitor and identify protein function and interactions responsible for the total activities of the cell. The application of these new proteomic technologies to study cellular activities, will lead to a faster sample throughput and increased sensitivity for the detection of individual proteins, thus providing major opportunities for the discovery of new biomarkers for the early detection of protein alterations associated with the progression of the disease state.

Use of proteomic patterns in serum to identify ovarian cancer

BACKGROUND

New technologies for the detection of early-stage ovarian cancer are urgently needed. Pathological changes within an organ might be reflected in proteomic patterns in serum. We developed a bioinformatics tool and used it to identify proteomic patterns in serum that distinguish neoplastic from non-neoplastic disease within the ovary.

METHODS

Proteomic spectra were generated by mass spectroscopy (surface-enhanced laser desorption and ionisation). A preliminary "training" set of spectra derived from analysis of serum from 50 unaffected women and 50 patients with ovarian cancer were analysed by an iterative searching algorithm that identified a proteomic pattern that completely discriminated cancer from non-cancer. The discovered pattern was then used to classify an independent set of 116 masked serum samples: 50 from women with ovarian cancer, and 66 from unaffected women or those with non-malignant disorders.

FINDINGS

The algorithm identified a cluster pattern that, in the training set, completely segregated cancer from non-cancer. The discriminatory pattern correctly identified all 50 ovarian cancer cases in the masked set, including all 18 stage I cases. Of the 66 cases of non-malignant disease, 63 were recognised as not cancer. This result yielded a sensitivity of 100% (95% CI 93--100), specificity of 95% (87--99), and positive predictive value of 94% (84--99).

INTERPRETATION

These findings justify a prospective population-based assessment of proteomic pattern technology as a screening tool for all stages of ovarian cancer in high-risk and general populations.