FEV_KEGG.Experiments.24 module

Context

Similar to 23, where group and super group were calculated with a fixed and a variable majority threshold. Maybe the group should be excluded from the super group to get less extreme results. The question answered would then be ‘EC numbers new to group, compared to ALL OTHER groups in super group’.

Question

Taking a complete set of organisms, forming a group (Brachyspirales, Order) and a super group (Spirochaetes, Class) of organisms, how many functions occur in the majority network of the group, but not in the majority network of the super group, excluding the organisms of the group? For a fixed and high majority threshold for the smaller group, but varying majority thresholds for the bigger super group.

Method

Similar to 21, only fixed majority threshold for group. - Create a group of all organisms of Order Brachyspirales to be found in KEGG. - Create a super group of all organisms of Class Spirochaetes to be found in KEGG, including the same organsims as the group of Order. - Exclude organisms of group from super group. - Calculate majority substance-ec graph for group. Leaving only EC numbers which occur in all organisms of the group. - REPEAT for varying majority-percentages: - Calculate majority substance-ec graph for super group. Leaving only EC numbers which occur in all organisms of the super group. - Calculate the difference of the two sets of majority EC numbers, leaving only the EC numbers which occur in Order majority, but not in Class majority. - Print the amount of EC numbers and their percentage of all EC numbers in Order, ie. how many of the EC numbers in Order do not exist in Class majority.

Result

300/328 -> 91.5% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 100% majority for super group and 80% majority for group
281/328 -> 85.7% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 90% majority for super group and 80% majority for group
271/328 -> 82.6% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 80% majority for super group and 80% majority for group
264/328 -> 80.5% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 70% majority for super group and 80% majority for group
254/328 -> 77.4% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 60% majority for super group and 80% majority for group
224/328 -> 68.3% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 50% majority for super group and 80% majority for group
203/328 -> 61.9% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 40% majority for super group and 80% majority for group
142/328 -> 43.3% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 30% majority for super group and 80% majority for group
 85/328 -> 25.9% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 20% majority for super group and 80% majority for group
 35/328 -> 10.7% of EC numbers in Brachyspirales are new, compared to Spirochaetes majority @ 10% majority for super group and 80% majority for group

Conclusion

Leaving out the group and comparing to the rest of the super group merely boosts the percentages for the lower thresholds. For higher thresholds, the group’s redundancy does not have a noticeable effect. If the group were bigger, though, the effect may well be noticeable and skew the higher thresholds, too. It should always be assured, that the super group is sufficiently larger than the group.