Limits...
ROS-DET: robust detector of switching mechanisms in gene expression.

Kayano M, Takigawa I, Shiga M, Tsuda K, Mamitsuka H - Nucleic Acids Res. (2011)

Bottom Line: Furthermore, for each of the top 10 pairs ranked by ROS-DET, we attempted to identify a pathway, i.e. consecutive biological phenomena, being related with the corresponding two genes by checking the biological literature.In 8 out of the 10 pairs, we found two parallel pathways, one of the two genes being in each of the two pathways and two pathways coming to (or starting with) the same gene.This indicates that two parallel pathways would be cooperatively used under one experimental condition, corresponding to the positive correlation, and the two pathways might be alternatively used under the other condition, corresponding to the negative correlation.

View Article: PubMed Central - PubMed

Affiliation: Bioinformatics Center, Institute for Chemical Research, Kyoto University, Gokasho, Uji 611-0011, Japan.

ABSTRACT
A switching mechanism in gene expression, where two genes are positively correlated in one condition and negatively correlated in the other condition, is a key to elucidating complex biological systems. There already exist methods for detecting switching mechanisms from microarrays. However, current approaches have problems under three real cases: outliers, expression values with a very small range and a small number of examples. ROS-DET overcomes these three problems, keeping the computational complexity of current approaches. We demonstrated that ROS-DET outperformed existing methods, under that all these three situations are considered. Furthermore, for each of the top 10 pairs ranked by ROS-DET, we attempted to identify a pathway, i.e. consecutive biological phenomena, being related with the corresponding two genes by checking the biological literature. In 8 out of the 10 pairs, we found two parallel pathways, one of the two genes being in each of the two pathways and two pathways coming to (or starting with) the same gene. This indicates that two parallel pathways would be cooperatively used under one experimental condition, corresponding to the positive correlation, and the two pathways might be alternatively used under the other condition, corresponding to the negative correlation. ROS-DET is available from http://www.bic.kyoto-u.ac.jp/pathway/kayano/ros-det.htm.

Show MeSH
Three types of negative cases of switching mechanisms.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3113587&req=5

Figure 2: Three types of negative cases of switching mechanisms.

Mentions: We then focus on the absolute difference of two correlation coefficients. The procedure of this approach is that we first compute the correlation coefficient of two genes in expression for each of the two experimental conditions and check the absolute difference of the two correlation coefficients. More concretely, in Figure 1, we first compute the correlation coefficient r1 between gene X and gene Y for one experimental condition, say class 1 (shown by +), and r2 between them for the other condition, say class 2 (shown by •). We then compute the absolute difference between these two correlation coefficients as the score of two genes X and Y as shown below:(1)A switching mechanism must have this of a larger value, because two correlation coefficients should be different in the switching mechanism. We can consider any measure of correlation coefficients in this approach, such as the Pearson's correlation coefficient (5,7) and the Spearman's rank correlation (6) [(3) is a review over differential coexpression, including the switching mechanism]. Because of various correlation coefficients, there can be many approaches of using the absolute difference of two correlation coefficients, but they still have problems by which negative cases can be detected as positives. The negatives which can be detected incorrectly are caused by the following three main reasons: (i) outliers, (ii) range bias and (iii) a small number of examples. Figure 2a illustrates a typical negative case of (i), where only two points of class 2 are in the upper-right area, which are outliers and make this case pretend to be a switching mechanism. Figure 2b shows a case of (ii), where the range of expression values of class 2 is much smaller than that by class 1. Regardless of the very small range, class 2 can show a strong negative correlation in its figure, by which the absolute difference between the two correlation coefficients would become large. In reality, however, the small range of class 2 makes us unconvincing whether class 2 is negatively correlated or not. Thus, we cannot say that Figure 2b is a switching mechanism, and this means that Figure 2b should be a negative case. Figure 2c shows a typical case of (iii), where the number of points in two classes is so small that it cannot be considered as a switching mechanism, implying that this cannot be a positive gene pair.Figure 2.


ROS-DET: robust detector of switching mechanisms in gene expression.

Kayano M, Takigawa I, Shiga M, Tsuda K, Mamitsuka H - Nucleic Acids Res. (2011)

Three types of negative cases of switching mechanisms.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3113587&req=5

Figure 2: Three types of negative cases of switching mechanisms.
Mentions: We then focus on the absolute difference of two correlation coefficients. The procedure of this approach is that we first compute the correlation coefficient of two genes in expression for each of the two experimental conditions and check the absolute difference of the two correlation coefficients. More concretely, in Figure 1, we first compute the correlation coefficient r1 between gene X and gene Y for one experimental condition, say class 1 (shown by +), and r2 between them for the other condition, say class 2 (shown by •). We then compute the absolute difference between these two correlation coefficients as the score of two genes X and Y as shown below:(1)A switching mechanism must have this of a larger value, because two correlation coefficients should be different in the switching mechanism. We can consider any measure of correlation coefficients in this approach, such as the Pearson's correlation coefficient (5,7) and the Spearman's rank correlation (6) [(3) is a review over differential coexpression, including the switching mechanism]. Because of various correlation coefficients, there can be many approaches of using the absolute difference of two correlation coefficients, but they still have problems by which negative cases can be detected as positives. The negatives which can be detected incorrectly are caused by the following three main reasons: (i) outliers, (ii) range bias and (iii) a small number of examples. Figure 2a illustrates a typical negative case of (i), where only two points of class 2 are in the upper-right area, which are outliers and make this case pretend to be a switching mechanism. Figure 2b shows a case of (ii), where the range of expression values of class 2 is much smaller than that by class 1. Regardless of the very small range, class 2 can show a strong negative correlation in its figure, by which the absolute difference between the two correlation coefficients would become large. In reality, however, the small range of class 2 makes us unconvincing whether class 2 is negatively correlated or not. Thus, we cannot say that Figure 2b is a switching mechanism, and this means that Figure 2b should be a negative case. Figure 2c shows a typical case of (iii), where the number of points in two classes is so small that it cannot be considered as a switching mechanism, implying that this cannot be a positive gene pair.Figure 2.

Bottom Line: Furthermore, for each of the top 10 pairs ranked by ROS-DET, we attempted to identify a pathway, i.e. consecutive biological phenomena, being related with the corresponding two genes by checking the biological literature.In 8 out of the 10 pairs, we found two parallel pathways, one of the two genes being in each of the two pathways and two pathways coming to (or starting with) the same gene.This indicates that two parallel pathways would be cooperatively used under one experimental condition, corresponding to the positive correlation, and the two pathways might be alternatively used under the other condition, corresponding to the negative correlation.

View Article: PubMed Central - PubMed

Affiliation: Bioinformatics Center, Institute for Chemical Research, Kyoto University, Gokasho, Uji 611-0011, Japan.

ABSTRACT
A switching mechanism in gene expression, where two genes are positively correlated in one condition and negatively correlated in the other condition, is a key to elucidating complex biological systems. There already exist methods for detecting switching mechanisms from microarrays. However, current approaches have problems under three real cases: outliers, expression values with a very small range and a small number of examples. ROS-DET overcomes these three problems, keeping the computational complexity of current approaches. We demonstrated that ROS-DET outperformed existing methods, under that all these three situations are considered. Furthermore, for each of the top 10 pairs ranked by ROS-DET, we attempted to identify a pathway, i.e. consecutive biological phenomena, being related with the corresponding two genes by checking the biological literature. In 8 out of the 10 pairs, we found two parallel pathways, one of the two genes being in each of the two pathways and two pathways coming to (or starting with) the same gene. This indicates that two parallel pathways would be cooperatively used under one experimental condition, corresponding to the positive correlation, and the two pathways might be alternatively used under the other condition, corresponding to the negative correlation. ROS-DET is available from http://www.bic.kyoto-u.ac.jp/pathway/kayano/ros-det.htm.

Show MeSH