PURPOSE and SCOPE: The bioinformatics service is a function of YCAS which serves as the Biostatistics Shared Resource for Yale Cancer Center (YCC). It is a highly interactive team of cancer biostatisticians who work collaboratively with basic, clinical, translational and population science researchers to advance the frontiers of cancer medicine and public health. Yale Cancer Center (YCC), in conjunction with the Yale Center for Analytical Sciences (YCAS), provides for the biostatistical needs of the entire YCC.
The services encompass bioinformatics support to YCC members on study designs, analysis, and grant and manuscript preparations. We aim to provide high-quality, cutting-edge and custom data analyses, as well as consultation and training/education for high throughput genomics, transcriptomics, proteomics, and other high-throughput data sets.
The best mechanism for engaging the BSR is through percent effort inclusion on grants. Absent this, there are limitations on the number of hours that can be dedicated to any individual project. The following guidelines will be used to prioritize the utilization of the BSR and to ensure fair utilization by the entire research community at YCC.
PRIORITIES: The priorities listed below will be used to triage and prioritize which studies will be eligible for BSR utilization. BSR resources are available only to members and associate members of the YCC for cancer related projects.
(a) Top priority is given to members with peer-reviewed funding, peer-reviewed grant applications, investigator-initiated protocols being developed from Yale discoveries, PRC-approved clinical trials, and trials under development. (b) Second priority is given to members with non-peer-reviewed grants and clinical studies. (c) Lower priority is assigned to non-investigator initiated industry sponsored trials, unfunded laboratory studies, and retrospective database studies.
These priorities reflect the maximum number of hours that the biostatisticians may be able to dedicate to any particular project without compensation.
- Array-based platforms, including genotyping arrays, gene expression arrays, copy number arrays, methylation arrays, and ChIP-on-chip arrays.
- Analysis of next-generation sequencing data, including:
- DNA-sequencing analysis, such as single nucleotide variations, indels, copy number variations, translocations and inversions from whole-exome and whole-genome DNA sequencing (DNA-seq)
- Gene expression, splice variants, gene-set, and pathway analysis from RNA sequencing (RNA-seq)
- microRNA expression from microRNA sequencing (miRNA-seq)
- DNA methylation and differential methylation identification from methylation sequencing (methyl-seq)
- Transcription factor binding sites and chromatin modifications from ChIP sequencing (ChIP-seq)
- RNA binding sites identification from CLIP sequencing (CLIP-seq)
- Proteomics data anyalsis
- Data annotation, visualization, and database integration
- User training in bioinformatics software and tools
Basic Science Support
- Experimental Design
- Power calculations
- Development of randomization schemes
- Basic analysis (e.g.; chi-square, t-test, regression, ANOVA)
- Graph generation
- Statistical protocol development
- Office hours
- Analytic and research and design clinics (to workshop ideas)
Clinical Studies Support
The following represents the priorities identified by YCC leadership:
- Investigator initiated clinical research trials with translational component
- Investigator initiated clinical trials
- Prospective IRB-approved translational trials in a protocol setting
- Cooperative group multi-center trials
- Industry sponsored trials
- Retrospective studies (laboratory based or otherwise) with clinical correlations
Population Science Support
The following represents the level of biostatistical support for population science:
- Study design, including sample size/power calculations and analytic plan
- Development of randomization schemes
- Grant and manuscript preparation
Efforts will be made to ensure parity of access for both basic, clinical, translational and population science projects. Nonetheless, limits on staffing may dictate that the scale of large jobs be capped. The size of such a cap will be determined by the BSR Oversight committee and is likely to change over time based on usage and staffing.
Currently, the guidelines for resources provided by YCC gratis will be as below. Additional resource hours will require funding by the investigator or by special approval by the BSR Committee through a special application. All investigators are entitled to receive a minimum of 16 hours per year. Limits are as below:
- Grant preparation for NIH funding – as many hours as reasonably required to prepare the grant. More than 40 hours requires committee approval.
- Unfunded analyses, including those for manuscript preparation – 16 hours total per investigator per year.
- Investigator-initiated trials–16 hours for study design; as many hours as reasonably required (to be reviewed by the Oversight Committee) to monitor, analyze and publish the study.
- IRB approved translational projects – 8 hours for design and 16 hours of analysis per project.
- Grant preparation for internal (e.g., TTARE) and external (non-NIH) funding – 8 hours.
- Analyses and manuscript preparation for peer-reviewed and funded grants- 8 hours (or percent effort) are per grant budget.
Data Management Support
In addition to the support outlined above, support for data management for basic, clinical, translational and population science will be available through the BSR on a fee for service basis.
Publication Credit and External Funding
Standards for academic credit vary among research disciplines. For small-scale routine consultations, BSR does not expect authorship.Reasonable expectations for larger projects must be guided by a sense of fairness and proportionality and community standards. The International Committee of Medical Journal Editors provides consensus guidelines on publication credit http://www.icmje.org/icmje-recommendations.pdf that can be used as a framework for publication credit discussions.
It is recommended that users of the resource have a direct discussion of these issues in the context of job specification discussions in early phases of the project so that a consensus can be achieved.
We provide supports for cancer-related projects. All Yale Cancer Center (YCC) members are entitled to receive free support for all non-billable services, and a maximum of 4 hours free support per year. Non-YCC members receive free support for non-billable activities, but need to pay for all billable activities. YCC members receive priority over nonmembers.
Non-billable activities: work on grant preparation, including analysis of pilot studies, provided the member of the YCAS Bioinformatics Core is written into the grant as percent FTE.
Billable activities: include all other services
Fees: $121/hour for FY15.
- The following estimation assumes no complications (e.g., quality issues) with the data. The actual analysis time heavily depends on data complexity.
- Turn-around time will depend on the work load.
- For sequencing data, the service is divided into two parts: hands-on time and computer time. The computer time is estimated with moderate coverage data based on an 8-CPU server. Actual computing time will vary from days to weeks depending on sequencing depth and server condition.
|gene/miRNA expression||Chip-on-chip||Methylation array|
|Array-based||4 hours/comparison||4 hours/sample||4 hours/sample|
|Whole-exome||Hands-on time vs. Computer time||2 hours/sample vs. 6-8 hours/sample||2 hours/sample vs. 3-4 hours/sample||2 hours/sample vs. 4 hours/sample|
|Whole-genome||Hands-on time vs. Computer time||3 hours/sample vs. 200 hours/sample||3 hours/sample vs. 100 hours/sample||3 hours/sample vs. 100 hours/sample|
|Mapping||Methylation calling||Differential methylation|
|Targeted (RRBS)||Hands-on time vs. Computer time||2 hours/sample vs. 6-8 hours/sample||2 hours/sample vs. 3-4 hours/sample||2 hours/sample vs. 4 hours/sample|
|Whole-genome||Hands-on time vs. Computer time||3 hours/sample vs. 200 hours/sample||3 hours/sample vs. 100 hours/sample||3 hours/sample vs. 20 hours/sample|
|Mapping||Gene expression||Pathway/gene-set||Splice analysis|
|Hands-on time||4 hours/comparison||2 hours/comaprison||1 hour/comparison||2 hours/sample|
|Computer time||1-2 days/comparison||10 hours/comparison||1 hour/comparison||4 hours/sample|
|Mapping||Peak calling||Differential binding analysis||Motif anaylsis|
|Hands-on time||2 hours/sample||2 hours/sample||2 hours/comparison||2 hours/sample|
|Computer time||5 hours/sample||2 hours/sample||4 hours/comparison||2 hours/sample|
|Mapping||Differential binding analysis||Motif analysis|
|Hands-on time||2 hours/sample||2 hours/comparison||2 hours/sample|
|Computer time||5 hours/sample||4 hours/comparison||2 hours/sample|
Protocol Review Committee: Marilyn Stolar, PhD - firstname.lastname@example.org
Cancer Prevention and Control Program: Ted Holford, PhD - email@example.com
Bioinformatics: Xiaoqing Yu, PhD - firstname.lastname@example.org
Administrative Support: Alicia Lakomski - (203) 737-5946 or email@example.com