QSAR model for pka prediction of phenols
Descriptors (topological, mathematical and quantum) were used to generate quantitative construction property connections (QSPR) for the pKa of 80 phenols. The informational index was divided into 56 preparation and 24 test sets, and models were built using the preparation set's incomplete least...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of Bologna
2023-02-01
|
| Series: | EQA |
| Subjects: | |
| Online Access: | https://eqa.unibo.it/article/view/15686 |
| Summary: | Descriptors (topological, mathematical and quantum) were used to generate quantitative construction property connections (QSPR) for the pKa of 80 phenols. The informational index was divided into 56 preparation and 24 test sets, and models were built using the preparation set's incomplete least squares (PLS) relapse. The consistency and predictive power of the best acquired QSAR models were achieved through internal approval, Y randomization, and external approval, and their pertinence area was confirmed by the influence technique. The benefits of the various direct relapse investigations' measurable boundaries. Standard deviation (S), standard deviation error of prediction (SDEP, External validation coefficient test), determination coefficient R², cross-validated R² (Q²) (SDEPext). The cross-validated R² (test Q²ext) values (95.68%, 95.22%, 0.304, 0.312, 0.292, and 96.24%, respectively) attest to the model's good fit. |
|---|---|
| ISSN: | 2039-9898 2281-4485 |