dsMTL Package

Development name

dsMTL

Name of server-side packages

dsMTLBase

Name of client-side packages

dsMTLClient

Date this information was late updated/checked

02/09/2021

Description of packages purpose

dsMTL (Federated Multi-Task Learning based on DataSHIELD) provided federated, privacy-preserving multi-task learning analysis. dsMTL aimed at simultaneously learning the outcome (e.g. diagnosis) associated patterns across datasets with dataset-specific, as well as shared, effects. Four sparse MTL methods were contained in dsMTL to disentangle specific structures of “shared-specific” components. In addition, Federated LASSO was also included due to its wide usage. dsMTL was suitable for biomedical applications, such as comorbidity analysis, multi-omics analysis, multi-modal data analysis, and high-dimensional molecular studies.

Image

How to contact developer institution/team/individual

Han Cao (hank9cao@gmail.com), Youcheng Zhang, Carl Herrmann and Emanuel Schwarz, Heidelberg University, Germany

Latest version

0.9

Type distribution licence

GNU General Public License v3.0

Methods of obtaining package

CRAN Address

-

What versions of R work with the package?

≥ 4.0.1

What R packages do the packages depend on?

dsMTLBase

Opal 3.0.3
dsBase 6.1.0
resourcer 1.0.1

dsMTLClient

DSI (≥ 1.2.0)
DSOpal (≥ 1.2.0)
corpcor
stats

Status

We are documenting the package

Is the package tested?
Yes
Is the package documented?
Not yet
Has the package had a disclosure audit?
No

Is the package suitable for deployment in the production environment? (Yes/No)

Yes

Does your package have features to protect the privacy of data, or does it just provide remote analysis functionality?

Yes, since multi-task learning outputted a model with multiple matrices i.e. a matrix capturing shared effect and matrices capturing cohort-specific effects. Therefore returning only the shared matrix was our strategy to defend the model inverse attack.

Additional Information