Table 1 Overview of existing libraries and applications. Majority of the approaches are open source with a focus on data protection against honest but-curious (HBC) adversaries. Federated protection is provided using combination of homomorphic encryption (HE), multiparty computation MPC), and differential privacy (DP). Majority of the libraries provide custom API for building applications but limited command line (CLI) tools that can be deployed immediately by users
From: COLLAGENE enables privacy-aware federated and collaborative genomic data analysis
Method | Link/reference | Description | Open Source | Application library | Adversary | Federated protection | Matrix functions | Custom CLI | Custom API |
|---|---|---|---|---|---|---|---|---|---|
SEAL | C++ implementation of RLWE-based HE schemes | Yes | Library | HBC | None | No | No | Yes | |
TenSEAL | Python wrapper for SEAL with focus on tensor processing | Yes | Library | HBC | None | No | No | Yes | |
HyFED | MPC-type privacy framework with Aggregator/Compensator entities for secret sharing | Yes | Library | HBC | MPC | No | No | Yes | |
MK-TFHE | C++ implementation of Multi-Key lattice-based HE | Yes | Library | HBC | HE | No | No | Yes | |
Lattigo | GO implementation of lattice-based multiparty HE schemes | Yes | Library | HBC | HE/MPC | No | No | Yes | |
PySyft | Python-based Federated machine learning with secure primitives based on SPDZ Protocol | Yes | Library | HBC | DP/MPC/HE | No | No | Yes | |
FedML | Python-based Federated machine learning | Yes | Library | HBC | None | No | No | Yes | |
TrustGWAS | Outsourcing for collaborative GWAS method using Asharov-type key sharing | No | Application | HBC | HE | No | No | No | |
Intel HE-Toolkit | SEAL/Palisade wrapper with Intel CPI Optimizations | Yes | Library | HBC | None | Yes | No | Yes | |
SEQURE | MPC-based custom python-like programming interface | Yes | Library | HBC | MPC | No | No | Yes | |
MPC GWAS | MPC-based GWAS method | Yes | Application | HBC | MPC over secure channel | No | No | No | |
sPLINK | Privacy-aware GWAS method via HyFED | Yes | Application | HBC | HyFed | No | No | No | |
OpenFL | Python-based Federation workflow | Yes | Library | HBC | None | No | No | Yes | |
Vantage6 | Python-based Federated Learning platform with UI | Yes | Application | HBC | None | No | Yes | Yes | |
Flower | Python-based Federated machine learning library | Yes | Library | HBC | MPC | No | No | Yes | |
FATE | Python-based Federated Learning platform with UI | Yes | Application | HBC | DP/MPC/HE | No | Yes | Yes | |
NVFlare | Python-based Federated machine learning library | Yes | Library | HBC | None | No | No | Yes | |
APPFL | Python-based Federated machine learning library | Yes | Library | HBC | None | No | No | Yes | |
COLLAGENE | C++ SEAL wrapper with threshold multiparty keys | Yes | Library | HBC | HE/MPC | Yes | Yes | Yes |