Preparing and Submitting RMIP study data to BDC – Phase 2: Data Management
C
Caryn Willis
started a topic
13 minutes ago
RMIP investigators are responsible for providing the scientific data for the data types outlined in the approved project plan to BDC. The NHLBI BDC Data Management Core (DMC) has developed a Data Management Plan that can be accessed on the RMIP Resource Hub that specifies data type-specific standards to enable the analysis of data across RMIP studies.
De-identification
All data should be de-identified. While the DMC can advise on strategies for deidentification and recommended data type specific tools, the responsibility lies with the investigator to ensure that participant privacy is protected.
File Naming
All files must start with a prefix that uses the following naming convention
The file name prefix consists of six components separated by underscores when applicable
The phrase “RMIP”
The project identifier is a 3-digit code that assigned to the study by the IDCCH
The participant ID is a unique deidentified identifier assigned to a participant (e.g., “001”, “A01”, “123456”) by the RMIP investigator and should remain constant across the project.
The discriminator is an alphabetic character assigned to a particular biospecimen collection event
The identifier is a numeric character assigned to a particular biospecimen collection event
The vial identifier is an alphabetic character assigned to a specific biospecimen collection aliquot
When you are ready to begin this step, please reach out to the DMC to ensure that study assigned identifiers are aligned with the linkers used for the molecular data. There may be instances where the study team generated data on samples that were not sent to the IDCCH for molecular profiling. Please reach out to the DMC so that a linker can be assigned for those files.
The combination of the sample discriminator and sample identifier uniquely identifies every collection event across all participants for the study. If the data were not produced directly from a biospecimen (e.g., MRI, echocardiogram) then the prefix would only include the first three components of the prefix.
Following the prefix, filenames should include a description of the data and terminate with the file extension. The format of the data description is flexible and left to the investigator to determine what is useful and appropriate. The filename should not include spaces or special characters. This table provides examples of appropriate filenames
Data Type
Example File Name
Whole Genome Sequencing from participant A1234. This was from the second biospecimen collected in the study and is from the third aliquot of the biospecimen
RMIP_001_A1234_B_002_C_WGS.fastq
RMIP_001_A1234_B_002_C_WGS_GRCh38.bam
Single cell RNA sequencing collected at baseline from participant012. This was from the sixth biospecimen collected in the study and is from the first aliquot of the biospecimen
RMIP_001_012_F_006_A_scRNA_baseline.fastq
RMIP_001_012_F_006_A_ scRNA_baseline_GRCh38.bam
Single cell RNA sequencing collected at 6 months from participant012. This was from the sixth biospecimen collected in the study and is from the first aliquot of the biospecimen
RMIP_001_012_F_006_A_scRNA_6m.fastq
RMIP_001_012_F_006_A_ scRNA_6m_GRCh38.bam
MRI imaging of the heart at baseline for participant 003
RMIP_009_003_MRI_baseline_heart.dicom
MRI imaging of the heart at 6 month follow up for participant 003
RMIP_009_003_MRI_6mo_heart_with_contrast.dicom
File Organization
The RMIP Data Management Plan specifies a standardize directory structure so that files are readily findable across all RMIP studies.
1st Level: Data on BDC are stored by consent group. There will be a separate cloud storage bucket for each consent group. Please separate the files into directories based on the consent group that the participant is in.
2nd Level: The data are then stratified into two sub-directories: “Raw” and “Harmonized”. “Raw data” is defined as data that would be provided to an investigator if the sample were to be processed at a core facility, rather than being defined as the typical output from the instrumentation. Harmonized data is defined as the data that follows the recommended data standards for content, structure, and file format.
3rd Level: The final level will split data by type. Directories will use the following naming convention:
The directory name consists of three components separated by an underscore.
The phrase “RMIP”
The project identifier is a 3-digit code that assigned to the study by the IDCCH.
The data type is the designated label for each data type being generated. Available data type codes can be found in Appendix A of the DMP (also available on the RMIP Resource Hub).
The name for longitudinal or repeated measure subdirectories should describe the collection circumstances (e.g., “RMIP_011_WGS _baseline”, “RMIP_011_WGS _6 months follow up”, “RMIP_011_WGS _replicate 1”).
If you are sharing a data type that is not included in Appendix A let us know and we can assign a label and revise the appendix.
If you are sharing a data type that is not included in Appendix A reach out to the DMC so that an appropriate label can be assigned and added to the appendix.
Data Dictionaries
A data dictionary describes the fields used in the various dataset files. This information in the data dictionary provides the context about how the data was collected to allow for proper interpretation of the data. The format of the data dictionary are provided in Appendix E of the data management plan and the DMC has prepared an MS Excel template for investigators.
Support
The DMC will continue to be available to provide support and guidance throughout the duration of this process. Please reach out if you have any questions to jstratford@rti.org or attend weekly BDC Office Hours on Wednesdays from 2:00-3:00PM ET.
Caryn Willis
RMIP investigators are responsible for providing the scientific data for the data types outlined in the approved project plan to BDC. The NHLBI BDC Data Management Core (DMC) has developed a Data Management Plan that can be accessed on the RMIP Resource Hub that specifies data type-specific standards to enable the analysis of data across RMIP studies.
De-identification
All data should be de-identified. While the DMC can advise on strategies for deidentification and recommended data type specific tools, the responsibility lies with the investigator to ensure that participant privacy is protected.
File Naming
All files must start with a prefix that uses the following naming convention
The file name prefix consists of six components separated by underscores when applicable
The phrase “RMIP”
The project identifier is a 3-digit code that assigned to the study by the IDCCH
The participant ID is a unique deidentified identifier assigned to a participant (e.g., “001”, “A01”, “123456”) by the RMIP investigator and should remain constant across the project.
The discriminator is an alphabetic character assigned to a particular biospecimen collection event
The identifier is a numeric character assigned to a particular biospecimen collection event
The vial identifier is an alphabetic character assigned to a specific biospecimen collection aliquot
When you are ready to begin this step, please reach out to the DMC to ensure that study assigned identifiers are aligned with the linkers used for the molecular data. There may be instances where the study team generated data on samples that were not sent to the IDCCH for molecular profiling. Please reach out to the DMC so that a linker can be assigned for those files.
The combination of the sample discriminator and sample identifier uniquely identifies every collection event across all participants for the study. If the data were not produced directly from a biospecimen (e.g., MRI, echocardiogram) then the prefix would only include the first three components of the prefix.
Following the prefix, filenames should include a description of the data and terminate with the file extension. The format of the data description is flexible and left to the investigator to determine what is useful and appropriate. The filename should not include spaces or special characters. This table provides examples of appropriate filenames
Data Type
Example File Name
Whole Genome Sequencing from participant A1234. This was from the second biospecimen collected in the study and is from the third aliquot of the biospecimen
RMIP_001_A1234_B_002_C_WGS.fastq
RMIP_001_A1234_B_002_C_WGS_GRCh38.bam
Single cell RNA sequencing collected at baseline from participant012. This was from the sixth biospecimen collected in the study and is from the first aliquot of the biospecimen
RMIP_001_012_F_006_A_scRNA_baseline.fastq
RMIP_001_012_F_006_A_ scRNA_baseline_GRCh38.bam
Single cell RNA sequencing collected at 6 months from participant012. This was from the sixth biospecimen collected in the study and is from the first aliquot of the biospecimen
RMIP_001_012_F_006_A_scRNA_6m.fastq
RMIP_001_012_F_006_A_ scRNA_6m_GRCh38.bam
MRI imaging of the heart at baseline for participant 003
RMIP_009_003_MRI_baseline_heart.dicom
MRI imaging of the heart at 6 month follow up for participant 003
RMIP_009_003_MRI_6mo_heart_with_contrast.dicom
File Organization
The RMIP Data Management Plan specifies a standardize directory structure so that files are readily findable across all RMIP studies.
1st Level: Data on BDC are stored by consent group. There will be a separate cloud storage bucket for each consent group. Please separate the files into directories based on the consent group that the participant is in.
2nd Level: The data are then stratified into two sub-directories: “Raw” and “Harmonized”. “Raw data” is defined as data that would be provided to an investigator if the sample were to be processed at a core facility, rather than being defined as the typical output from the instrumentation. Harmonized data is defined as the data that follows the recommended data standards for content, structure, and file format.
3rd Level: The final level will split data by type. Directories will use the following naming convention:
The directory name consists of three components separated by an underscore.
The phrase “RMIP”
The project identifier is a 3-digit code that assigned to the study by the IDCCH.
The data type is the designated label for each data type being generated. Available data type codes can be found in Appendix A of the DMP (also available on the RMIP Resource Hub).
The name for longitudinal or repeated measure subdirectories should describe the collection circumstances (e.g., “RMIP_011_WGS _baseline”, “RMIP_011_WGS _6 months follow up”, “RMIP_011_WGS _replicate 1”).
If you are sharing a data type that is not included in Appendix A let us know and we can assign a label and revise the appendix.
If you are sharing a data type that is not included in Appendix A reach out to the DMC so that an appropriate label can be assigned and added to the appendix.
Data Dictionaries
A data dictionary describes the fields used in the various dataset files. This information in the data dictionary provides the context about how the data was collected to allow for proper interpretation of the data. The format of the data dictionary are provided in Appendix E of the data management plan and the DMC has prepared an MS Excel template for investigators.
Support
The DMC will continue to be available to provide support and guidance throughout the duration of this process. Please reach out if you have any questions to jstratford@rti.org or attend weekly BDC Office Hours on Wednesdays from 2:00-3:00PM ET.
Additional Resources
NHLBI Instructions for Preparing Clinical Research Study Datasets for Submission
NHLBI Data Sharing FAQs
Best practices for clinical trials data harmonization and sharing on NHLBI BDC, learned from CONNECTS network COVID-19 studies