Data analysis apparatus, data analysis method, method of generating trained model, system, and program

1 . A data analysis apparatus comprising:

a processor; and

a storage having recorded thereon a trained model machine-trained to output, when the trained model receives input of an FT-IR spectrum and first information designating one atomic group included in a trained atomic group, second information indicating whether the inputted FT-IR spectrum includes a peak derived from the atomic group designated by the inputted first information, wherein

the trained atomic group includes an atomic group containing at least three atoms,

the processor is configured to

request a user to select at least one atomic group from among options including the atomic group containing at least three atoms,

determine whether the at least one atomic group selected by the user includes the atomic group containing at least three atoms, and

based upon the determination that the at least one atomic group selected by the user includes the atomic group containing at least three atoms, input an analysis target which is the FT-IR spectrum and the first information that designates the atomic group containing at least three atoms selected by the user to the trained model to analyze the analysis target, and

the processor is configured not to analyze the analysis target by the trained model when the processor determines the at least one atomic group selected by the user does not include the atomic group containing at least three atoms.

2 . The data analysis apparatus according to claim 1 , wherein

the processor is configured to

obtain, for each of the at least one atomic group selected in response to the request, a result of analysis indicating whether the peak derived from the atomic group is included in the analysis target, and

control a notification apparatus to notify the user of the result of analysis.

3 . The data analysis apparatus according to claim 2 , the storage further having recorded thereon analysis software that analyzes, when the analysis software is given an FT-IR spectrum, whether the given FT-IR spectrum includes a peak derived from a prescribed atomic group on a rule basis, wherein

the options further include an atomic group containing less than three atoms that is included in the prescribed atomic group but not included in the trained atomic group, and

the processor is configured to

obtain with the trained model, when the atomic group containing at least three atoms is included in the at least one atomic group selected in response to the request, the result of analysis associated with the selected atomic group containing at least three atoms, and

obtain with the analysis software, when the atomic group containing less than three atoms is included in the at least one atomic group selected in response to the request, the result of analysis associated with the selected atomic group containing less than three atoms.

4 . The data analysis apparatus according to claim 3 , wherein

the trained atomic group includes at least one atomic group selected from the group consisting of a nitro group, an ester group, an acrylic acid structure, a toluene structure, a thiophene structure, a pyrimidine structure, a toluidine structure, an aromatic carboxylic acid structure, a benzamide structure, a salicylic acid structure, a benzimidazole structure, and a benzothiazole structure, and

the prescribed atomic group includes at least one atomic group selected from the group consisting of a fluoro group, a chloro group, a bromo group, an iodo group, a hydroxy group, an amino group, and a cyano group.

5 . The data analysis apparatus according to claim 3 , wherein

the processor is configured to, in response to analysis of the analysis target being completed for all atomic group selected by the user, control the notification apparatus to display the analysis target and the result of analysis for each atomic group selected by the user, the result of analysis indicating whether a peak derived from the atomic group is included in the analysis target.

6 . The data analysis apparatus according to claim 2 , wherein

the processor is configured to

control the notification apparatus to display a first input screen for receiving input by the user, of a storage location where the analysis target is stored,

obtain the analysis target from the storage location inputted by the user, and

control the notification apparatus to display a second input screen for receiving input of the at least one atomic group among the options by the user, the second input screen showing a message that requests the user to select the at least one atomic group from among the options.

7 . A data analysis method comprising:

obtaining an FT-IR spectrum of a compound having an unknown molecular structure;

requesting a user to select at least one atomic group from among options including the atomic group containing at least three atoms;

determining whether the selected at least one atomic group includes an atomic group containing at least three atoms; and

based upon the determination that the selected at least one atomic group includes the atomic group containing at least three atoms, inputting the obtained FT-IR spectrum and designation information that designates the atomic group containing at least three atoms selected by the user to a trained model, wherein

the trained model is a mathematical model machine-trained to output, when the trained model receives input of an FT-IR spectrum and the designation information that designates the atomic group containing at least three atoms, information indicating whether the inputted FT-IR spectrum includes a peak derived from an atomic group containing at least three atoms designated by the inputted designation information.

8 . A system including at least one computer that performs the method according to claim 7 .

9 . A non-transitory computer readable medium storing a program, wherein

the program causes a computer to perform the method according to claim 7 .

10 . A method of generating a trained model comprising:

analyzing, by a Fourier transform infrared spectrophotometer including an analysis apparatus and a control device, a compound having an already known molecular structure with a Fourier transform infrared spectrophotometer to obtain an FT-IR spectrum;

storing, by the control device, the obtained FT-IR spectrum and atomic group information of the analyzed compound in association with each other; and

carrying out, by a server configured to communicate with the control device, machine learning for generating a trained model capable of analyzing whether an FT-IR spectrum includes a peak derived from a prescribed atomic group, with the stored FT-IR spectrum and the stored atomic group information being used as training data, wherein

the prescribed atomic group does not include an atomic group containing less than three atoms and includes an atomic group containing at least three atoms, and

the atomic group information of a compound including the prescribed atomic group indicates the prescribed atomic group included in the compound, and the atomic group information of a compound not including the prescribed atomic group indicates that the compound does not include the prescribed atomic group.