Non-targeted metabolomics_Wuhan biological Optical Valley Science and technology

Login| Register| 中文

搜索



Message

Telephone

027-5981 1720

WeChat

TOP

Current location：

Homepage

/

Technology Services

/

Small molecule metabolite detection services

/

Metabolomics service

/

Non-targeted metabolomics

技术平台 



Single Cell Sequencing Service 
- Single Cell Sequencing 
- Spatial transcriptome sequencing 
  - Spatial Gene Expression for Fresh Frozen
  - Spatial Gene Expression for FFPE
High throughput sequencing services 
Cell-based immune-related assays 
Proteomics testing service 
- Proteomics services by LC/MS 
- Olink Proteomics Services 
  - Olink Precision Proteomics Analysis Service
Small molecule metabolite detection services 
- Metabolomics service 
  - Targeted metabolomics
  - Non-targeted metabolomics
- Nucleic acid mass spectrometry detection service

Non-targeted metabolomics

Product Description

Case Analysis

Result display

Sample delivery

Q&A

Non-targeted metabolomics (i.e. discovery metabolomics) is a non-hypothetical approach that aims to obtain as many metabolites as possible from a single analysis. The main purpose is to compare the metabolomes (all metabolites of a given organism) of the control and experimental groups to find out the differences in their metabolites and to explore the metabolic pathways between the differential metabolites. The analysis generally includes:

a. Metabolic profiling (also called differential expression analysis): looking for metabolites of interest with statistically significant abundance changes in a set of experimental and control samples.

b. Identification: After performing metabolic profiling, the chemical structure of these metabolites is determined.

c. Interpretation: The final step in the research process, explaining the association between the metabolites found and the biological process or biological state.

Research Process

The non-targeted metabolomics research process can be mainly divided into experimental design, sample collection and processing, metabolite extraction and concentration, sample detection, data analysis, metabolite identification, and finally biological interpretation. Because the metabolome changes extremely fast, the metabolite species are various, the concentration varies greatly, the chemical properties are different, and the data information is huge, each step may have a large impact on the final results, so it is crucial to realize the standardized operation of metabolomics

Technology platform

The main platforms for non-targeted metabolomics are nuclear magnetic resonance (NMR), gas chromatography-mass spectrometry (GC-MS) and liquid chromatography-mass spectrometry (LC-MS), each of which has a certain bias, so researchers need to choose the appropriate platform according to the purpose of the study and the type of samples. The purpose is to achieve a more complete and comprehensive study.

Table Comparison of major assay platforms for non-targeted

Major Assay Platforms	Advantages	Disadvantages
NMR	1. High throughput 2. Universality 3. Good objectivity and reproducibility 4. Fast (2~3 min/sample) 5. Simple pretreatment, no derivatization and separation required 6. Non-invasive 7. Can detect most organic compounds in the sample 8. Long instrument life	1. Detection dynamic range is narrow 2. Lower sensitivity and resolution than MS 3. The sample amount is relatively large (0.1 ~ 0.5 mL) 4. Expensive instrumentation and maintenance costs
GC-MS	1. High throughput 2. High precision, sensitivity and reproducibility 3. Has a reference standard spectral database, easy to characterize 4. The sample amount is moderate (0.1~0.2 mL) 5. Can detect most organic and some inorganic molecules in the sample	1. Requires derivatization 2. Requires separation 3. Slow analysis speed (20~40 min/sample) 4. It is difficult to identify new compounds 5. Not suitable for the analysis of difficult volatile, thermally unstable substances
LC-MS	1. High throughput 2. High separation rate and sensitivity 3. Wide dynamic range of detection 4. Simple sample processing, no need for derivatization 5. Small sample size (10~100 μL) 6. Applicable to thermally unstable, non-volatile, not easily derivatized and large molecular weight substances	1. Slow analysis speed (15~40 min/sample) 2. Lack of standard spectral database for reference 3. Difficult to identify new compounds 4. Higher cost 5. Short instrument life

Data analysis methods

The data obtained from non-targeted metabolomics are complex and multidimensional, and the chemometric methods need to be fully applied to deeply explore the information in them. The data are generally pre-processed first, and the specific process includes normalization, data transformation, centralization, and normalization steps. Only after pre-processing, simplification and dimensionality reduction of the data are realized, and reliable data models are established. The pattern recognition methods applied to metabolomics data analysis mainly include unsupervised learning methods and supervised learning methods.

The unsupervised learning methods mainly include: principal component analysis (PCA), nonlinear mapping, cluster analysis, etc. Supervised learning methods are mainly based on partial least squares (PLS), neural network improvement methods, least squares-discriminant analysis (PLS-DA), orthogonal least squares (OPLS), etc., among which PCA and PLS-DA are commonly used pattern recognition methods in metabolomics.

Non-targeted metabolomics data processing process

Sample types

Serum - Plasma - Urine - Cerebrospinal fluid - Saliva - Feces - Microorganisms - Cells - Plants - Tissues – Other

Lead time

　　The non-targeted metabolomics services and period offered by Wuhan Biobank are listed in the following table.：

Platform	Instruments	Database	Number of compounds detected	Cycle time	Detection price
GC-MS	Aglient 7890A/5975C	1、NIST 2、Fiehn RTX5 3、Self-built database	Tens to hundreds	About 15 working days	Please inquire
UPLC-TOF	SCIEX TripleTOF 5600 plus	1、Self-built database 2、METLIN 3、HMDB 4、KEGG	thousands	About 20 working days	Please inquire

Case Study

Article: Mathé EA, et al., Noninvasive urinary metabolomic profiling identifies diagnostic and prognostic markers in lungcancer. 2014, Cancer Research, 74(12):3259-70 (IF=9.32)

Experimental material

Pre-screening sample (1005 cases): sample collection process over a decade (1998-2007): 469 patients (pre-treatment urine) + 536 healthy individuals

Post validation samples (158 cases): recent samples collected (2008-2010): 80 patients (pre-treatment urine) + 78 healthy individuals

Tissue samples: 48 cases of tumor tissues and their surrounding non-tumor tissues

Procedure

Overview: pre-targeted non-targeted screening + post-targeted MRM validation

Experimental process:

Conclusion

Pre-screening: In the non-targeted metabolomic analysis of pre-screening samples, a total of 1807 signals were detected in the positive ion mode and 1359 in the negative ion mode. The detection of smoking-related metabolites (Cotinine, Nicotine-N'-oxide and Trans-3'-hydroxycotinine) revealed that the relevant indicators could well distinguish the smoking population from the non-smoking population, which validated the feasibility of this analysis method.

Data analysis

a. Diagnostic and typing study: After applying statistical analysis to exclude the interference of ethnicity and gender, the authors identified four differential metabolites: NANA, Cortisol sulfate, Creatine riboside and 561+ (unidentified substance). ROC analysis revealed that the AUC values of the four differential metabolites ranged from 0.63 to 0.76 in all populations, while in the lung cancer population of stage 1 - 2, the AUC values ranged from 0.59 to 0.70. Using Creatine riboside alone, or using all four metabolites for prediction gave more accurate results (p<0.00001).

　　Overlap between signals that are predictive of lung cancer status in the training set based on the Random Forests classification.

　　Receiver Operating Characteristic (ROC) analysis of individual metabolites and their combination

b. Prognostic study: after taking into account various factors such as gender, ethnicity, disease stage, tissue section, smoking history, radiotherapy, and surgery, the authors found that high levels of NANA, Cortisol sulfate, Creatine riboside, and 561+ would result in a poor prognostic outcome (low survival), and that the four metabolites and survival correlation, although independent, had a cumulative effect; high levels of Creatine riboside and 561+ also reduced patient survival in the stage I-II lung cancer population.

　　Kaplan-Meier survival estimates in the training set are depicted for the top four predictive metabolites

c. Post validation: Again using a non-targeted metabolomics approach, Creatine riboside, NANA and 561+ were found to be significantly elevated in the urine of patients in the validation population (80 patients + 78 healthy individuals). The relevance of these four metabolites to the disease was validated by selecting 198 samples (92 patients + 106 healthy individuals) for targeted metabolomics after excluding age, gender and ethnicity interference. In addition, to demonstrate the stability of these four metabolites, the authors kept the samples for two years and then repeated the experimental results again to prove their value for clinical application.

　　Abundance and validation of metabolites that were top contributors in the classification of patients as lung cancer or healthy controls.

d. Extension of the mechanistic study: the authors finally examined the levels of the above four metabolites in tumor tissues and found that the levels of Creatine riboside and NANA were significantly higher in tumor tissues compared to neighboring tissues, while Creatine levels were also increased.

　　Linking urinary metabolites to lung cancer tissue metabolome.

Summary.

This article is a very classic example of metabolomic biomarker study, which contains pre-targeted screening plus post-targeted validation, and is well worth learning.

The article finally returns to tissue samples from body fluid samples, opening the door to later mechanistic studies. It is also worthwhile to learn from this approach to clinical research.

　　Sample delivery instructions

Assay strategy/method	Blood samples	Urine	Cerebrospinal fluid	Stool	Microorganisms	Cells	Tissue	Plant
Non-targeted metabolome	500ul	1ml	200ul	2g	250mg	107个	250mg	250mg
Targeted metabolome	500ul	1ml	200ul	2g	250mg	107个	250mg	250mg
NMR metabolome	500ul	1ml	500ul	2g	500mg	107个	500mg	500mg

2. Reproducibility requirements of metabolomics delivery biology

The differential changes in the metabolome are the presentation of the amplification effect of genomic changes, and the volatility of metabolites in different individuals also becomes larger. On the other hand, differential analysis in metabolomics is generally based on multivariate statistical analysis methods such as PCA and PLS-DA, and only the principal components sampled with larger sample sizes are representative of the population. Therefore, compared to technical methods such as genomics, metabolomics requires more biological replicates of the samples.

a. Clinical specimens greater than 30 cases.

b. greater than 10 cases for model animal samples.

c. Plant microbial samples greater than 8 cases per group.

d. Cellular samples greater than 6 cases.

3.Metabolomics send sample collection criteria

Sample type		Collection criteria
Blood	Serum	3~4mL of blood is drawn intravenously and placed in a blood collection tube (without any anticoagulant or procoagulant); left at 4℃ for 30~60min (no more than 2h) to clot; centrifuged at 3500rpm, 4℃ for 10 minutes, and after the blood cells have completely settled, the upper layer of serum is aspirated and stored at -80℃.
	Plasma	Blood samples were collected and placed in collection tubes containing appropriate amounts of anticoagulant (Citrate anticoagulant was used with caution; Heprin anticoagulant was recommended for NMR; EDTA anticoagulant (K2, K3, Na, absolutely no Li) was recommended for mass spectrometry; the collection tubes were repeatedly inverted to mix the anticoagulant with the blood; the blood samples were centrifuged at 2000-3000 rpm for 10 min (within 1 h of collection at room temperature). Centrifugation at 2000-3000 rpm for 10 min (within 1 h at room temperature after collection; blood samples can be processed within 4 h on ice after collection), and the supernatant is plasma; stored in a -80°C refrigerator.
Urine		Collect urine (do not add sodium azide or other bacteriostatic agents to avoid interference with mass spectrometry) in a centrifuge tube (PP material recommended), centrifuge at 10,000 rpm for 10 minutes at 4°C, and store the supernatant at -80°C. It is recommended that the urine creatinine test be performed at the same time as the metabolomics test.
Other body fluid samples		The samples should be collected without adding other reagents and then frozen in liquid nitrogen after centrifugation and stored at -80°C.
Cells	Adherent cells (not less than 1.0E7 cells per dish)	Before processing, count the cells in each dish, discard the culture medium in the dish, add 10ml of deionized water to each dish to wash off the medium, shake for 2s (not too vigorously to prevent cell rupture), add 15ml of liquid nitrogen to quench (2mins), add 1.5ml of methanol pre-chilled at -80℃, and scrape with a spatula.
	Suspended cells	Transfer the complete suspension to a 1.5 ml centrifuge tube, centrifuge for 10 min at 4 °C at 13000 rpm, and take a fixed volume of supernatant (equal volume per dish, as much as possible, e.g. 0.8 mL) in a 1.5 mL centrifuge tube and store at -80 °C. At the same time, the cell debris was weighed and recorded.
	Cell culture medium (2mL/case)	A fixed volume of cell culture fluid was aspirated from each dish to be tested and placed in a 2mL centrifuge tube for 10 minutes at 10000rpm, 4°C. The supernatant was removed and stored at -80°C.
Feces		Pick 3 spots from different parts of the feces, weigh and place in a centrifuge tube (do not add sodium azide or other bacteriostatic agents to avoid interference with mass spectrometry), store at -80°C.
Tissue	Animal tissue samples	Operate on ice throughout the entire procedure to minimize handling time. Remove blood (heart, liver, kidney and lung can be perfused, other tissues can be washed several times with pre-cooled PBS), snap frozen in liquid nitrogen and stored at -80°C.
	Human tissue samples and tumor samples	Minimize handling time. Remove blood (pre-cooled PBS dripping several times), snap-freeze in liquid nitrogen, and store at -80°C.
Microbiological and bacterial samples		Collect 10^7 bacteria or 500 mg/bacterial filament in centrifuge tubes. If the specific amount of mycelium cannot be judged, the size of a soybean grain can be used as a standard, rapidly snap frozen in liquid nitrogen and stored at -80℃.
Plant samples		It is recommended to crush and weigh, liquid nitrogen flash freeze and store at -80℃. Uncrushed samples can also be sent.