RECORD
An Inverse QSAR Method Based on Linear Regression and Integer Programming
Machine Learning · Submitted 6 Jul 2021
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim The results of computational experiments suggest our method can infer chemical graphs with around up to 50 non-hydrogen atoms.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Recently a novel framework has been proposed for designing the molecular structure of chemical compounds using both artificial neural networks (ANNs) and mixed integer linear programming (MILP). In the framework, we first define a feature vector $f(C)$ of a chemical graph $C$ and construct an ANN that maps $x=f(C)$ to a predicted value $\eta(x)$ of a chemical property $\pi$ to $C$. After this, we formulate an MILP that simulates the computation process of $f(C)$ from $C$ and that of $\eta(x)$ from $x$. Given a target value $y^*$ of the chemical property $\pi$, we infer a chemical graph $C^\dagger$ such that $\eta(f(C^\dagger))=y^*$ by solving the MILP. In this paper, we use linear regression to construct a prediction function $\eta$ instead of ANNs. For this, we derive an MILP formulation that simulates the computation process of a prediction function by linear regression. The results of computational experiments suggest our method can infer chemical graphs with around up to 50 non-hydrogen atoms.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Recently a novel framework has been proposed for designing the molecular structure of chemical compounds using both artificial neural networks (ANNs) and mixed integer linear programming (MILP). | background 0.99 | 0.06 | 0.04 | asserted 1.00 |
| s1 | In the framework, we first define a feature vector $f(C)$ of a chemical graph $C$ and construct an ANN that maps $x=f(C)$ to a predicted value $\eta(x)$ of a chemical property $\pi$ to $C$. | method 0.85 | 0.40 | 0.03 | asserted 0.99 |
| s2 | After this, we formulate an MILP that simulates the computation process of $f(C)$ from $C$ and that of $\eta(x)$ from $x$. | method 0.89 | 0.37 | 0.03 | asserted 0.99 |
| s3 | Given a target value $y^*$ of the chemical property $\pi$, we infer a chemical graph $C^\dagger$ such that $\eta(f(C^\dagger))=y^*$ by solving the MILP. | method 0.78 | 0.34 | 0.08 | asserted 0.99 |
| s4 | In this paper, we use linear regression to construct a prediction function $\eta$ instead of ANNs. | contribution 0.58 | 0.93 | 0.09 | asserted 1.00 |
| s5 | For this, we derive an MILP formulation that simulates the computation process of a prediction function by linear regression. | method 0.86 | 0.95 | 0.07 | asserted 1.00 |
| s6 | The results of computational experiments suggest our method can infer chemical graphs with around up to 50 non-hydrogen atoms. | result 1.00 | 0.93 | 0.58 | hedged possibility 1.00 |
What was examined, and what was not
- arXiv listing abstract v3
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Recently a novel framework has been proposed for designing the molecular structure of chemical compounds using both artificial neural networks (ANNs) and mixed integer linear programming (MILP). In the framework, we first define a feature vector $f(C)$ of a chemical graph $C$ and construct an ANN that maps $x=f(C)$ to a predicted value $\eta(x)$ of a chemical property $\pi$ to $C$. After this, we formulate an MILP that simulates the computation process of $f(C)$ from $C$ and that of $\eta(x)$ from $x$. Given a target value $y^*$ of the chemical property $\pi$, we infer a chemical graph $C^\dagger$ such that $\eta(f(C^\dagger))=y^*$ by solving the MILP. In this paper, we use linear regression to construct a prediction function $\eta$ instead of ANNs. For this, we derive an MILP formulation that simulates the computation process of a prediction function by linear regression. The results of computational experiments suggest our method can infer chemical graphs with around up to 50 non-hydrogen atoms.
Human review
How the literature uses it
1 arXiv paper cite this in 6 sentences
Papers citing it · 5 sentences without a confident label
Uses its method or tool 1 paper
-
We employ the modeling of chemical compounds that was introduced by Zhu et al. [ref] .
Cycle-Configuration: A Novel Graph-theoretic Descriptor Set for Molecular Inference · Modeling of Chemical Compounds · 1.00
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.