The paper aims to analyse and propose scientific methods suitable for evaluating the adaptation quality of Virtual Learning Environments (VLEs) matching Informatics learners' needs. The authors' approach consists of the consecutive application of the principles of multiple criteria decision analysis for identifying the VLEs adaptation quality criteria, sets portrait method to analyse the interconnections of the VLE adaptation quality criteria and the learners' computational thinking skills, fuzzy group decision making theory to obtain final evaluation measures of the VLEs quality criteria, and scalarization method to obtain the final results of evaluating the VLEs quality. While applying these methods, appropriate decision support system was developed. This system consists of the learners' computational thinking skills' questionnaires, observations results and conclusions, VLEs adaptation quality criteria, their ratings (values) and weights, and final evaluation results that propose a proper decision. This approach should help Universities and schools to create, buy, or find free VLE software mostly suitable for teaching and learning Informatics. Computational thinking term is detailed in the paper, and interconnected with the VLEs adaptation quality criteria using sets portrait method. After that, multiple criteria decision analysis approach is used to evaluate the adaptation quality of VLE in terms of its conformance with the learners' computational thinking styles. The experts' additive utility function is proposed to use for the expert evaluation of the adaptation quality of VLEs. Trapezoidal fuzzy numbers method is proposed to use for establishing both weights and ratings (values) of the VLEs quality criteria matching learners' computational thinking styles. Practical example of the experimental evaluation of three popular open source VLEs is also presented in the paper. Presented research results are particularly useful for Informatics/software engineering education.