InSysBio published LikelihoodProfiler package for Identifiability Analysis

February 8, 2019

February 8, 2019

InSysBio, a pioneer in Quantitative Systems Pharmacology (QSP), modeling and simulation for drug development, announced a publication of LikelihoodProfiler package in the official Julia repository.

LikelihoodProvider is a software package for practical Identifiability Analysis. The reliability and predictability of a kinetic systems biology (SB) and systems pharmacology (SP) model depends on the calibration of model parameters. Taking into account the lacking of data and the experimental variability the value of any parameter determined unambiguously. This results in characterization of parameter by “confidence intervals” or even “non-identifiable” parameters when the confidence interval is open. The package includes algorithms to perform practical identifiability analysis and evaluation confidence intervals using LikelihoodProfiler. Results of the identifiability analysis can be used to qualify and calibrate parameters or to reduce the model.

Evgeny Metelkin, Head of Development Department at InSysBio, explains:

“The package includes a number of algorithms for practical identifiability analysis. Along with linear and quadratic extrapolation algorithms we introduce CICO (Confidence Intervals by Constrained Optimization) algorithm which was presented at ICSB 2018. All the algorithms can be effectively applied for wide range of models and data type because they do not require calculation of gradients in opposite to the previous methods. CICO algorithm can be efficiently applied to complex kinetic models where function differentiability is not guaranteed and each likelihood point is computationally expensive.”

The package is based on Julia programming language and distributed for free under the MIT license.

About InSysBio
InSysBio is a Quantitative Systems Pharmacology (QSP) company located in Moscow, Russia (INSYSBIO LLC) and Edinburgh, UK (INSYSBIO UK LIMITED). InSysBio was founded in 2004 and has an extensive track record of helping pharmaceutical companies to make right decisions on the critical stages of drug research and development by application of QSP modeling. InSysBio’s innovative approach has already become a part of the drug development process implemented by our strategic partners: nowadays there are more than 100 completed projects in collaboration with leaders of pharmaceutical industry. The company has published multiple scientific studies in various therapeutic areas. InSysBio team is developing software and tools for QSP and continuously improves methods for biological modeling. For more information about InSysBio, its solutions and services, visit