Highly self-defined natural product duplicate removal and annotation method

Abstract

The invention relates to a highly self-defined natural product duplicate removal and annotation method, which comprises the following steps of: 1, fermenting a target strain under different fermentation conditions, and performing data acquisition on fermentation liquor in a high-resolution liquid chromatography-mass spectrometry non-targeted mode; 2, processing the collected original data to obtain a signal feature list only appearing in the strain fermentation liquor; and 3, performing MNA analysis on the obtained signal feature list to obtain annotations and clustering results. According to the method, the annotation efficiency of the high-flux mass spectrum data is improved, time-consuming and labor-consuming repeated labor in the natural product discovery process can be avoided, and the discovery efficiency of novel active natural products is improved.

Description

Technical Field

[0001] The present invention relates to a highly customized natural product deduplication and annotation method, Molecular Networking Annotation (MNA), which belongs to the field of natural product chemistry. Background Art

[0002] In the process of discovering active natural products, high performance liquid chromatography and mass spectrometry can often provide rich information for complex samples, but the large amount of secondary mass spectrometry data obtained in high throughput in the non-targeted acquisition mode of liquid chromatography-mass spectrometry is difficult to obtain. How to effectively annotate and remove duplicates from the massive data obtained is the key to obtaining new natural products. Summary of the invention

[0003] In view of the above situation, in order to overcome the shortcomings and deficiencies of the prior art, the present invention provides a highly customized natural product deduplication and annotation method.

[0004] The present invention is based on a rapid deduplication and annotation strategy of high-resolution liquid chromatography-mass spectrometry. Taking the fermentation broth of strains under different culture medium fermentation conditions as an example, by collecting multi-stage mass spectra of fermentation broth and blank culture medium, multi-stage mass spectrometry comparison is used to obtain database annotation results and clustering within samples, novel natural products in the samples are selected, and rapid annotation and deduplication of high-throughput mass spectrometry data are achieved.

[0005] The objective of the present invention is achieved through the following technical solutions:

[0006] A highly customized natural product deduplication and annotation method, specifically:

[0007] S1. Use multiple culture media to ferment the target strain to obtain different fermentation broths, and use the same culture medium as a blank control;

[0008] S2. After processing various fermentation broths and empty culture medium of the strain, use the data-independent mass spectrometry acquisition mode to collect the primary high-resolution mass spectrometry signal and its secondary mass spectrometry signal to obtain the mass spectrometry raw data, pre-process it into an open data format file that is easy to operate, and operate the open data format file. The main steps are feature extraction, ion flow construction, feature alignment and subtraction of blank culture medium signals. Finally, a list of mass spectrometry signal features that only appear in the fermentation broth of the target strain is screened out as the input of MNA;

[0009] S3. Analyze the features of the filtered mass spectrometry signals, and first compare these features with the database: first perform a primary mass spectrometry comparison, and then perform a secondary mass spectrometry comparison on those above the threshold according to the set threshold. The annotation results are retained according to the set threshold. Then, perform a secondary mass spectrometry comparison on the filtered signals, cluster them according to the set threshold, and divide them into clusters and single points. Finally, the clustering and annotation results are obtained.

[0010] Furthermore, in step S2, the collected primary and secondary high-resolution mass spectrometry signal raw data are processed into an open data format, and after feature extraction, ion flow construction, feature alignment and blank culture medium subtraction, a list of signal features that only appear in the fermentation broth of the target strain is finally screened out and used as the output of the MNA, specifically:

[0011] S21. Use software to convert the obtained mass spectrum raw file into an open data format;

[0012] S22. Use software to process open data format mass spectrometry files, extract primary mass spectrometry signals according to specified thresholds, extract secondary mass spectrometry signals according to specified thresholds, construct ion flow according to specified parameters, perform feature alignment as specified and subtract blank culture medium signals, and finally output a list of signal features that only appear in the fermentation broth of the target strain.

[0013] Further, the mass spectrometry signal characteristics after screening are compared with the database as described in step S3: firstly, the primary mass spectrometry is compared, and the secondary mass spectrometry is compared for those above the threshold according to the set threshold, and the annotation results are retained according to the set threshold; the secondary mass spectrometry is continued to be compared for the screened signals, and clustering is performed according to the set threshold, and divided into clusters and single points, and finally the clustering and annotation results are obtained, specifically:

[0014] S31. Based on a given error range, taking the signal list features appearing in the fermentation broth of all target strains as the object, all the primary mass spectra thereof are compared with the database, and the matching results within the error range are retained;

[0015] S32. further comparing the secondary mass spectra according to the primary mass spectrum matching results within the retained error range, and retaining the annotation results based on the similarity threshold;

[0016] S33. Compare the secondary mass spectra of all signal features appearing in the fermentation broth of the target strain in pairs, and divide the features into clusters or single points according to a given threshold.

[0017] Furthermore, the present invention also includes: S4 obtaining a product signal that can be further separated and purified based on the annotation and clustering of the signal molecules retained in the product signal of the target strain.

[0018] The present invention adopts a variety of fermentation conditions to maximize the efficiency of strains in metabolizing natural products; through a hierarchical comparison strategy of multi-level mass spectrometry, the candidate range is first narrowed by a primary mass spectrometry comparison, and then a secondary mass spectrometry comparison with a higher time complexity is used, which greatly improves the comparison efficiency; the mass spectrometry features in the sample are divided and clustered by the secondary mass spectrometry comparison, which can effectively expand the annotation information and enhance the deduplication effect. In addition, the most common natural product annotation process currently has the disadvantages of few optional parameters and a single mass spectrometry comparison method. However, the data of LC-MS is affected by many factors such as experimental samples and mass spectrometers, and the mass spectrometry data generated are different, so it is particularly important to select appropriate comparison parameters and comparison methods. The present invention relates to a highly customized natural product deduplication and annotation algorithm, which provides a variety of parameters and mass spectrometry comparison algorithms for selection, and supports the rapid adjustment of various parameters and methods. In summary, the new high-resolution LC-MS data interpretation strategy is of great significance for mining novel natural products from high-throughput data.

[0019] The advantages of the present invention are: improving the annotation efficiency of high-throughput mass spectrometry data, being able to avoid time-consuming and labor-intensive repetitive work in the natural product discovery process, and thus improving the discovery efficiency of novel active natural products. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 Flow chart of the method of the present invention.

[0022] Figure 2 Specific annotation and clustering results of crude extracts of strain K. viridogrisea DSM 43850 under different culture conditions.

[0023] Figure 3 The present invention provides specific annotations and clustering results for the crude extract of strain K. viridogrisea DSM 43850 under different culture conditions.

[0024] Figure 4 Schematic diagram of the structure of an aromatic glycoside compound cluster additionally annotated by MNA. DETAILED DESCRIPTION

[0025] The present invention ferments a target strain using multiple culture media to obtain different fermentation broths, and uses the same culture media as a blank control; after processing the multiple fermentation broths of the strain and the empty culture media, a primary high-resolution mass spectrometry signal and a secondary mass spectrometry signal thereof are collected using a data-independent mass spectrometry acquisition mode to obtain mass spectrometry raw data, which are pre-processed into an open data format file that is easy to operate; the open data format file is operated, and the main steps include feature extraction, ion flow construction, feature alignment and deduction of blank culture medium signals; finally, a list of mass spectrometry signal features that only appear in the fermentation broth of the target strain is screened out as an input of an MNA; the screened mass spectrometry signal features are analyzed as an object, and these features are first compared with a database: first, a primary mass spectrometry comparison is performed, and a secondary mass spectrometry comparison is performed for those above the threshold according to a set threshold, and annotation results are retained according to the set threshold, and then a secondary mass spectrometry comparison is performed for the screened signals, and clustering is performed according to the set threshold to divide them into clusters and single points, and finally clustering and annotation results are obtained to determine subsequent experimental tracking signals.

[0026] The present invention is further described in detail below in conjunction with examples and drawings, but the implementation method of the present invention is not limited thereto. The materials involved in the following examples can be obtained from commercial channels unless otherwise specified. The methods described are conventional methods unless otherwise specified.

[0027] The target strain Kutzneria viridogrisea JCM3282 used in the fermentation was purchased from Japan Type Culture Collection.

[0028] A highly customized natural product deduplication and annotation algorithm of the present invention, such as Figure 1 As shown, it specifically includes the following steps:

[0029] (1) The target strain is fermented under different culture medium conditions, and the blank culture medium is fermented under the same conditions, and samples are prepared for use, which specifically includes the following sub-steps:

[0030] (1.1) Target strains fermented in different culture media

[0031] Take 100uL of K. viridogrisea frozen glycerol and inoculate it into 100mL LB medium (10g Tryptone, 5gyeast extract, 10g NaCl, 1000mL distilled water) and activate it at 30℃ 180rpm for two days. Then, 7 kinds of 100mL liquid culture media were inoculated with 1% inoculum, as follows: GSM (4g glucose, 4g yeast extract, 10g maltextract, 1000mL distilled water), R2A (0.5g yeast extract, 0.5g difco proteosepeptone No3, 0.5g casamino acids, 0.5g glucose, 0.5g soluble starch, 0.3g sodium pyruvate, 0.3g K2HPO4, 0.05g MgSO4·7H2O, 1000mL distilled water), MHB (2g beef infusion solids, 1.5g starch, 17.5g casein hydrolysate, 1000mL distilled water), LB medium, 2216E (5g Bacto peptone, 1g Bacto yeast extract, 0.1g Fe(III) citrate, 19.45g NaCl, 5.9g MgCl2(anhydrous),3.24g Na2SO4,1.8g CaCl2,0.55g KCl,0.16gNaHCO3,0.08g KBr,34mg SrCl2,22mg H3BO3,4mg Na-silicate,2.4mg NaF,1.6mg(NH4)NO3,8mg Na2HPO4,1000mL Distilled water), TSB (17g casein peptone, 3g soypeptone, 2.5g D(+)-glucose, 5g NaCl, 2.5g K2HPO4, 1000mL distilled water), M42 (2gyeast extract, 10g soluble starch, 1000mL distilled water), fermented at 30°C and 180rpm for 7 days.

[0032] (1.2) After fermentation, add 10% volume of XAD-16N macroporous resin to each bottle, extract at 180rpm for 30min, elute with methanol, and spin dry to obtain crude extract. Take 1mg of each crude extract and dissolve it in 1mL of methanol, centrifuge at 13000rpm for 15min, and take the supernatant for later use.

[0033] (2) After the fermentation broth is prepared for use, data is collected in a non-targeted mode using high-resolution liquid chromatography-mass spectrometry, the raw data is processed into an open data format, and preprocessed into an MNA input format, including the following sub-steps:

[0034] (2.1) LC-MS data acquisition

[0035] LC-MS conditions were as follows: column, Agilent poroshell 120SB-C18 50×2.1 mm, 1.7 μm; mobile phase, (A) water and (B) methanol; flow rate, 0.3 mL / min; injection volume, 1 μL; gradient, 0-2 min (10% B), 2-18 min (10-100% B), 18-22 min (100% B), 22-25 min (10% B). Column temperature, 30 °C. MS was set up as X500B (Sciex AB) as follows: duration, 25 min; scan duration: 0.495; estimated cycle time, 3030; ion source, TurbolonSpray; curtain gas, 35 psi, ion source gas 1, 35 psi, ion source gas 2, 35 psi; temperature, 600 °C; acquisition mode, information-dependent acquisition; polarity, positive; spray voltage, 5500 V. TOF MS settings: TOF start mass, 200Da; TOF stop mass 2000Da; accumulation time, 0.2s; declustering voltage, 80V; collision energy, 10V; ID Acriteria.

[0036] (2.2) LC-MS data processing

[0037] The collected raw data were converted into '.mzml' format by MSconvert and preprocessed using MZmine2.5.1. The main processing steps included MS1 feature detection, ADAP chromatogram builder, Localminimum search, join aligner, and finally a list of signal features that only appeared in the fermentation broth of the target strain was screened out and used in '.mgf' and '.csv' file formats as input for MNA.

[0038] (3) Compare the filtered mass spectrometry signal features with the database and compare the features within the sample pairwise, and finally obtain the clustering and annotation results, including the following steps

[0039] (3.1) Setting parameters

[0040] MNA analysis uses the following default parameters: PEPMASS MATCH TOLERANCE, 10ppm; LIBRARY MATCHINGMETHOD, modified cosine; SELF CLUSTERING METHOD, neutral loss; SELF CLUSTERINGSIMILARITY, 0.7; SELF CLUSTERING PEAKS, 5; TOP K, 1; PEAKS,5;LIBRARY MATCHING SIMILARITY,0.7;LIBRARY MATCHING PEAKS,5;PEAK PERCENTAGE THRESHOLD,0.7.

[0041] (3.2) Database comparison

[0042] Based on the given error range, the primary mass spectra of the screened signal features are compared with the database (in this example, 10 ppm is set as the allowable error). For the comparison results with an error less than the set value, their secondary mass spectra are further compared (in this example, 0.7 is set as the secondary mass spectrum similarity threshold, 5 is the minimum number of matching peaks allowed for the secondary mass spectrum, and topK is 1, indicating that the result with the highest similarity is retained). The annotation results that meet the custom parameters are retained.

[0043] (3.3) Pairwise comparison of features within samples

[0044] Based on the given error range, the secondary mass spectra of all signal features appearing in the fermentation broth of the target strain were compared pairwise (in this example, 0.7 was set as the secondary mass spectrometry similarity threshold, and 5 was set as the minimum number of matching peaks allowed by the secondary mass spectrometry), and the features were divided into clusters or single points according to the given threshold.

[0045] After the analysis is completed, the molecular network annotation results are obtained ( Figure 3 ), add additional annotations using 7 culture medium information ( Figure 2 ), which can help quickly identify the source of features and also obtain more annotation results than traditional FBMN ( Figure 2). The MNA algorithm additionally annotated an unknown compound cluster containing novel aromatic glycoside compounds that could not be annotated by the traditional FBMN using a single secondary mass spectrometry similarity algorithm ( Figure 4 ), is a group of new natural products that can be further prepared and isolated.

[0046] Obviously, the embodiments described above are only descriptions of the preferred modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should fall within the protection scope determined by the claims of the present invention.

Claims

1. A highly customized natural product deduplication and annotation method, characterized in that: The steps include: S1. Use multiple culture media to ferment the target strain to obtain different fermentation broths, and use the same culture medium as a blank control; S2. After processing various fermentation broths and empty culture medium of the strain, use the non-data-dependent mass spectrometry acquisition mode to collect the primary high-resolution mass spectrometry signal and its secondary mass spectrometry signal to obtain the mass spectrometry raw data, pre-process it into an open data format file that is easy to operate, and operate the open data format file, including the following steps: feature extraction, ion flow construction, feature alignment and subtraction of blank culture medium signal, and finally screen out the mass spectrometry signal feature list that only appears in the fermentation broth of the target strain as the input of MNA; S3. Analyze the filtered mass spectrometry signal features as the object, first compare the filtered mass spectrometry signal features with the database: first perform a primary mass spectrometry comparison, and then perform a secondary mass spectrometry comparison on those above the threshold according to the set threshold, and retain the annotation results according to the set threshold; continue to perform a secondary mass spectrometry comparison on the filtered signals, cluster them according to the set threshold, and divide them into clusters and single points, and finally obtain the clustering and annotation results.

2. The method according to claim 1, characterized in that In step S2, the collected primary and secondary high-resolution mass spectrometry signal raw data are processed into an open data format, and after feature extraction, ion flow construction, feature alignment and blank culture medium deduction, a list of signal features that only appear in the fermentation broth of the target strain is finally screened out and used as the output of the MNA, specifically: S21. Use software to convert the obtained mass spectrum raw file into an open data format; S22. Use software to process open data format mass spectrometry files, extract primary mass spectrometry signals according to specified thresholds, extract secondary mass spectrometry signals according to specified thresholds, construct ion flow according to specified parameters, perform feature alignment as specified and subtract blank culture medium signals, and finally output a list of signal features that only appear in the fermentation broth of the target strain.

3. The method according to claim 1, characterized in that , the mass spectrometry signal characteristics after screening described in step S3 are compared with the database: firstly, the primary mass spectrometry is compared, and the secondary mass spectrometry is compared for those above the threshold according to the set threshold, and the annotation results are retained according to the set threshold; the secondary mass spectrometry is continued to be compared for the screened signals, and clustering is performed according to the set threshold, and divided into clusters and single points, and finally the clustering and annotation results are obtained, which are specifically: S31. Based on a given error range, taking the signal list features appearing in the fermentation broth of all target strains as the object, all the primary mass spectra thereof are compared with the database, and the matching results within the error range are retained; S32. further comparing the secondary mass spectra according to the primary mass spectrum matching results within the retained error range, and retaining the annotation results based on the similarity threshold; S33. Compare the secondary mass spectra of all signal features appearing in the fermentation broth of the target strain in pairs, and divide the features into clusters or single points according to a given threshold.

4. The method according to claim 1, characterized in that , also includes: Step S4 obtains the product signal that can be further separated and purified according to the annotation and clustering of the signal molecules retained in the product signal of the target strain.

Images