Our Operating Policy:

1. No patient identifiable experimental data can be generated as the outcome of the analysis


2. No human identifiable primary data, including genomic data, can reside on any partition shared on the cluster


3. Storage and preprocessing of primary data from human subjects should be performed in an isolated and secure computing and database environment


4. Users are urged to encrypt their de-identified genetic data in the scratch partitions as human identity can potentially be revealed through similarity analyses
• Please see our GPG and OpenSSL tutorial on how to secure your data


5. Before moving data to the cluster partition, users must make sure that:
• Human imaging data is pre-processed by removing facial features or using skull-stripped images
• Human genomic data is de-identified by imputation and re-coding of genomic data to strip the primary genotypes
• Even though the primary genotypes are stripped, the potential for data matching remains based on the statistical match versus reference genotype
• Therefore, users must adhere to strong individual user security policies