How many pathways in kegg

GHOSTX: an improved sequence homology search algorithm using a query suffix array and a database suffix array. PLoS One. Chemical and genomic evolution of enzyme-catalyzed reaction networks. FEBS Lett. Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Sign In. Advanced Search.

Search Menu. Article Navigation. Close mobile search navigation Article Navigation. Volume Article Contents Abstract. KEGG: new perspectives on genomes, pathways, diseases and drugs. Minoru Kanehisa , Minoru Kanehisa.

Oxford Academic. Miho Furumichi. Mao Tanabe. Yoko Sato. Kanae Morishima. Select Format Select format. Permissions Icon Permissions. Table 1. The KEGG databases. Database name. KEGG identifier. Open in new tab. Table 2. URL form. Table 3. Sequence data. Primary data source. Gene identifier a. Table 4. Architecture of KEGG website. Top pages KEGG home www. Table 5. Subject-oriented entry points to KEGG. Open in new tab Download slide.

Table 6. Manually drawn KEGG reference pathway maps. Number of maps a. Metabolism Global map 4 Overview map 5 Regular map Chemical structure transformation map 9 Genetic information processing Regular map 22 Environmental information processing Regular map 38 Cellular processes Regular map 24 Organismal systems Regular map 78 Human diseases Regular map 81 Drug development Drug structure map Table 7. KEGG Mapper tools. Query dataset. Search Pathway KOs, gene identifiers, C numbers, etc.

Google Scholar Crossref. Search ADS. For commercial re-use, please contact journals. Issue Section:. Download all slides. Comments 0. Add comment Close comment form modal.

I agree to the terms and conditions. You must accept the terms and conditions. Add comment Cancel. Table 1. Table 2. Table 3. Output format The output of all operations is in a text format: tab-delimited text returned from list, find, conv and link flat file database format returned from get text message returned from info.

The HTTP status code can be used to check if the operation was successful. Code Meaning Success Bad request syntax error, wrong database name, etc. Name info — display database release information and linked db information. This operation displays the database release information with statistics for the databases shown in Table 1.

Except for kegg, genes and ligand, this operation also displays the list of linked databases that can be used in the link operation. Current statistics in the KEGG website. Name list — obtain a list of entry identifiers and associated definition. This operation can be used to obtain a list of all entries in each database. First, the significant pathways can be identified through the DC cut-off calculated by formula 3 with the given significance level. To the given simulated data, when the significance level p -value was set at 0.

But when the significance level was set at 0. This result demonstrated that more significant pathways can be identified with the significance level increasing. Second, the direct and indirect determination factors from the DC can clearly display the correlated regulation among the pathways x 1 , x 2 , x 3 and x 4. However, the indirect regulations of x 1 , x 2 and x 3 to x 4 were all negative. The larger negative indirect regulation led to the negative DC value. The detailed comparison of indirect regulation determination of x 1 , x 2 and x 3 to x 4 showed that the negative regulation of pathway x 3 was the largest and that of pathway x 2 was the smallest.

Third, the sign of DC value can be used to predict the impact direction of pathways. The calculated results of simulated data demonstrated that the impact direction of pathways x 1 and x 2 were up-regulated and that of the pathways x 3 and x 4 were down-regulated.

We have also compared the DC value R j of decision analysis and the total effect r jy of path analysis [ 28 ] based on the simulated data Table 2. The path analysis demonstrated the identification of the significant pathways and the regulation among pathways through the total effect r jy and its subdivision. In fact, the regulations between pathways were mutual and non-equivalent.

The decision coefficient of decision analysis gave consideration to the regulation and retro-regulation of pathway x j on the basis of the subdivision of the coefficient of determination of path analysis. For example, pathway x 2 had a lower rank fourth according to the total effect of path analysis. In contrast, pathway x 2 had a higher rank first according to the DC value of decision analysis due to the retro-regulation of pathway x 2 to x 1 , x 3 and x 4.

The strategy of borrowing information from retro-regulation allows the decision analysis to identify the most significant and mainly contributed pathways. It is true that there is no gold standard to compare the methods in real studies because the biological truth is unknown. Therefore, the analysis results based on the simulated data only help to illustrate the distinctive characteristics of decision analysis. The DIA impact values of the KEGG pathways from the functional analysis of the bovine mammary transcriptome during the lactation cycle were chosen to test the utility of decision analysis model [ 30 ].

The decision analyses results of the other pathway categories were also attached in Additional file 1 : Table S8. The detailed impact data of selected KEGG pathway categories and subcategories from to vs. In addition, few pathways were deleted in that the number of missing data of these pathways was greater than or equal to three. Meanwhile, when the number of the missing data included in the pathway was less than three, they were filled with the average value of the other values belonging to this pathway.

The filled data were marked in red color in Additional file 2 : Table S1. In order to compare the results of impact direction produced by the decision analysis model and the DIA method, the detailed impact direction data of selected KEGG pathway categories and subcategories from to vs.

Similarly, the pathways including the missing data were processed as mentioned above. The most impacted pathways identified according to different DC cutoff values were displayed in Additional file 4 : Table S3. The detailed comparison results of all pathways under the decision analysis model and DIA method were displayed in Additional file 6 : Table S5. The decision trees of selected pathway categories and subcategories were displayed in Additional file 8 : Figure S1 according to the decision percentage.

According to the path analysis approach, the total CD R 2 of the selected KEGG pathway categories and subcategories had been calculated. These results showed that the observed outcomes were replicated by the model very well. The detailed ratios of direct and indirect CD for all selected pathways were shown in Table 3.

Similarly, the indirect CD ratios of almost all subcategory pathways were greater than their corresponding direct CD ratios. In short, the fact that almost all indirect CD ratios were greater than their corresponding direct CD ratios further revealed that the complex regulating mechanisms existed and were very important in the KEGG pathways.

In order to use a more suitable DC cut-off to identify the most impacted pathways, the significance levels of 0.

The results showed that the different DC cut-offs were identified for different category and subcategory pathways.

After integration, three DC cut-offs 0. It should be noted that the cut-off of 0. Therefore, when the absolute value of calculated DC for a pathway was greater than or equal to 0. Four subcategories are found to be the most inhibited pathways. In some cases, the most impacted pathways highlighted by the decision analysis model match our expectations. It is well known that the three main components of milk in dairy cow are lactose, fat and protein [ 29 ]. In other cases, the pathways are not immediately expected, but subsequent investigations revealed that these pathways identified by decision analysis are supported by previous experiment results.

In fact, the glycosphingolipid synthesized by these pathways have been reported to display beneficial health properties, especially for the defense of newborns against pathogens [ 36 ]. In addition, gangliosides have an important role in membrane function including cell signaling, cell adhesion and protein sorting [ 37 ].

In still other cases, no direct corroborative evidence could be found e. Thus, this finding serves as a hypothesis for future testing. The details of compare results are listed in Additional file 5 : Table S4. The results demonstrated that the concordance rate of pathway impact direction under decision analysis was significantly higher than that under KEGG-PATH approach when DIA pathway impact directions were used as standard. From the view of all the secondary pathways belonging to the same category, the concordance rates of pathway impact direction were also obviously improved from Based on this comparison, several distinctions between the three approaches can be made.

First, overwhelming majority of the most relevant function pathways in the mammary gland during lactation are captured based on DC values Additional file 5 : Table S4. The results showed that the correlation regulations strengthen the direct determination of these secondary pathways to some extent.

These results also can be confirmed by the subdivision of decision coefficient. Additional file 7 : Table S6 b In addition, the decision analysis method highlights some more biologically meaningful results. This result is potentially the most interesting given the strong literature support described above.

It is interesting that the inhibition of gangliosides presents in the results. This is in consistent with the fact that the concentration of glycosphingolipids showed a large decrease during the transition from colostrums to mature milk [ 38 ]. Second, the pathways were identified as the most impacted pathways based on the DC values, but they were not found according to the mean DIA impact value. These subdivision results revealed that the correlation regulation among pathways highlights the importance of these two subcategories.

On the contrary, this subcategory has the smallest average impact value in DIA approach. The phenomenon showed that the correlation regulation was very important in identifying the most impacted pathways.

The result can be supported by the fact that many of the products of these two pathways could be precursors of TCA cycle pathway [ 30 ]. But their DIA mean impact values are relatively small. The result indicated that the correlation regulation has resulted in the change of the importance of these pathways. The result demonstrated that the retro-regulation among these pathways should be very important.

Therefore, researchers should pay much more attention to these correlation regulations. To shed light on the difference, we checked the subdivision of decision coefficient. The reduction of fatty acid metabolism also can be supported by the fact that the fatty acid taken up by the mammary tissue mainly was used towards the synthesis of milk fat, including the components of cellular membranes [ 30 , 39 ]. Conversely, few of the most impacted pathways based on mean DIA impact values were not found according to DC values.

As Additional file 9 : Table S7 a showed that the two subcategories have very small direct determination and were slightly regulated by the other pathways. The importance of these two subcategories was weakened just because of the approximate balance of direct and indirect determination. Third, the results based on the decision analysis model were displayed through the construction of decision tree Additional file 8 : Figure S1.

Meanwhile, the red and black numbers were used to denote the decision percentage of KEGG subcategory pathway to its corresponding category pathway, the decision percentage of the secondary KEGG pathway to its corresponding subcategory pathway, respectively.

In this way, the researchers can catch important information fleetly and exactly. The first superiority of the decision analysis model was that the retro-regulation of each pathway was considered in identifying the most significant pathways based on the coefficient of determination.

Thus, the two pathways had the largely negative decision-making ability. Another superiority of the decision analysis model was that the impact directions of pathways could be estimated preliminarily and directly according to the sign positive or negative of the DC. Still further, the sign of the DC also gave consideration to the dependences among the pathways.

These results match our expectations because the lipid metabolism had a lot to do with the synthesis of the lactose and the reduction of fatty acid metabolism was considered towards the synthesis of milk fat through taking up the fatty acids by the mammary tissue. Besides, the fact that the impact direction of the TGF-beta pathway was negative based on the decision analysis was in accordance with the fact that this pathway appeared to have a negative role on mammary cell proliferation [ 40 ].

In this study, a decision analysis model is first proposed to identify the most impacted pathways. The decision analysis model borrows the decision coefficient to judge the importance of the pathways, which not only considers the direct determination factor of pathway itself, but also adds the correlation indirect determination factor with the other related pathways.

Compared with DIA approach, the decision analysis method overcomes the deficiency of analyzing each pathway independently. Compared with KEGG-PATH approach, the decision analysis method constructs a DC index based on the coefficient of determination of regression analysis, rather than correlation coefficient.

Importantly, the retro-regulation among pathways was considered in decision analysis. Therefore, the decision analysis model is a statistical data mining at a deeper level.

For the estimation of impact direction, the DIA method averages the impact direction values of the pathway during different time course. However, the decision analysis can judge the impact direction directly through the sign of decision coefficient. More importantly, the sign of decision coefficient was caused by the correlated regulation from the other related pathways.

Thus, the identification of pathway impact direction up-regulating or down-regulating through the decision coefficient also gave consideration to the dependences among the pathways. In addition, the regulation mechanisms among pathways can be demonstrated through the subdivision of decision coefficient. This numerical expression of the correlation regulation among pathways is another major highlight of the decision analysis model. The construction of decision tree can visually display the results of decision analysis.

We have developed a program in Matlab Ra, version 7. In the calculations, we found that the results might be inaccurate when the correlation matrix was close to singular or badly scaled. Although the decision analysis model is designed to analyze the KEGG pathways, it is theoretically also applicable to the other databases with similar dependency structure, such as Reactome, Wikipathways, etc.

However, considering the information about how cell and tissue type, age, and environmental exposures affect pathway interactions, how to apply the decision analysis to general cases with the original gene expression value rather than the DIA impact values is still a challenge.