Limits...
SIRAC: Supervised Identification of Regions of Aberration in aCGH datasets.

Lai C, Horlings HM, van de Vijver MJ, van Beers EH, Nederlof PM, Wessels LF, Reinders MJ - BMC Bioinformatics (2007)

Bottom Line: We first determine the DNA-probes that are important to distinguish the classes of interest, and then evaluate in a systematic and robust scheme if these relevant DNA-probes are closely located, i.e. form a region of amplification/deletion.SIRAC does not need any preprocessing of the aCGH datasets, and requires only few, intuitive parameters.The results on two breast cancer datasets show promising outcomes that are in agreement with previous findings, but SIRAC better pinpoints the dissimilarities between the classes of interest.

View Article: PubMed Central - HTML - PubMed

Affiliation: Bioinformatics group, Delft University, Delft, The Netherlands. c.lai@tudelft.nl

ABSTRACT

Background: Array comparative genome hybridization (aCGH) provides information about genomic aberrations. Alterations in the DNA copy number may cause the cell to malfunction, leading to cancer. Therefore, the identification of DNA amplifications or deletions across tumors may reveal key genes involved in cancer and improve our understanding of the underlying biological processes associated with the disease.

Results: We propose a supervised algorithm for the analysis of aCGH data and the identification of regions of chromosomal alteration (SIRAC). We first determine the DNA-probes that are important to distinguish the classes of interest, and then evaluate in a systematic and robust scheme if these relevant DNA-probes are closely located, i.e. form a region of amplification/deletion. SIRAC does not need any preprocessing of the aCGH datasets, and requires only few, intuitive parameters.

Conclusion: We illustrate the features of the algorithm with the use of a simple artificial dataset. The results on two breast cancer datasets show promising outcomes that are in agreement with previous findings, but SIRAC better pinpoints the dissimilarities between the classes of interest.

Show MeSH

Related in: MedlinePlus

Artificial dataset, sensitivity and specificity for s ∈ {2, 9}. Sensitivity, specificity and False Positive Rate (FPR) for two values of the parameter s, i.e. s ∈ {2, 9}. For each plot, on the horizontal axis are the different amplification lengths u, used, and on the vertical axis are the different amplitudes of the amplification m. The colors code the value of the sensitivity, specificity and FPR from 0 to 1.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2198922&req=5

Figure 3: Artificial dataset, sensitivity and specificity for s ∈ {2, 9}. Sensitivity, specificity and False Positive Rate (FPR) for two values of the parameter s, i.e. s ∈ {2, 9}. For each plot, on the horizontal axis are the different amplification lengths u, used, and on the vertical axis are the different amplitudes of the amplification m. The colors code the value of the sensitivity, specificity and FPR from 0 to 1.

Mentions: In our algorithm, we combine the different window sizes in order to obtain a unique region of amplification, by setting the parameter s. A location is amplified if it is judged amplified in s window sizes. We also investigated the effect of the parameter s. The top four plots of Figure 3 illustrates the sensitivity and specificity for two values of the parameter s, i.e. s ∈ {2, 9}. We choose s = 2 as a loose constraint, while the more strict value of s = 9 requires the consensus of two-thirds of the window sizes. For each plot, the horizontal axis depicts the different amplification lengths, u used, and the vertical axis the amplitudes of the amplification, m. The colors code the value of the sensitivity and specificity from 0 to 1. The small amplification of m = 0.2 is very difficult to detect, therefore the sensitivity is very low regardless of the length of the amplification (bottom row of blue squares in Figure 3(a)). When the amplification amplitude increases, the sensitivity rises as well. If s = 9 fewer extremely large and small aberrations are not detected compared to s = 2, in other words, the sensitivity is lower when s = 9 compared to s = 2. However, at the same time, the specificity increases (Figure 3(d)).


SIRAC: Supervised Identification of Regions of Aberration in aCGH datasets.

Lai C, Horlings HM, van de Vijver MJ, van Beers EH, Nederlof PM, Wessels LF, Reinders MJ - BMC Bioinformatics (2007)

Artificial dataset, sensitivity and specificity for s ∈ {2, 9}. Sensitivity, specificity and False Positive Rate (FPR) for two values of the parameter s, i.e. s ∈ {2, 9}. For each plot, on the horizontal axis are the different amplification lengths u, used, and on the vertical axis are the different amplitudes of the amplification m. The colors code the value of the sensitivity, specificity and FPR from 0 to 1.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2198922&req=5

Figure 3: Artificial dataset, sensitivity and specificity for s ∈ {2, 9}. Sensitivity, specificity and False Positive Rate (FPR) for two values of the parameter s, i.e. s ∈ {2, 9}. For each plot, on the horizontal axis are the different amplification lengths u, used, and on the vertical axis are the different amplitudes of the amplification m. The colors code the value of the sensitivity, specificity and FPR from 0 to 1.
Mentions: In our algorithm, we combine the different window sizes in order to obtain a unique region of amplification, by setting the parameter s. A location is amplified if it is judged amplified in s window sizes. We also investigated the effect of the parameter s. The top four plots of Figure 3 illustrates the sensitivity and specificity for two values of the parameter s, i.e. s ∈ {2, 9}. We choose s = 2 as a loose constraint, while the more strict value of s = 9 requires the consensus of two-thirds of the window sizes. For each plot, the horizontal axis depicts the different amplification lengths, u used, and the vertical axis the amplitudes of the amplification, m. The colors code the value of the sensitivity and specificity from 0 to 1. The small amplification of m = 0.2 is very difficult to detect, therefore the sensitivity is very low regardless of the length of the amplification (bottom row of blue squares in Figure 3(a)). When the amplification amplitude increases, the sensitivity rises as well. If s = 9 fewer extremely large and small aberrations are not detected compared to s = 2, in other words, the sensitivity is lower when s = 9 compared to s = 2. However, at the same time, the specificity increases (Figure 3(d)).

Bottom Line: We first determine the DNA-probes that are important to distinguish the classes of interest, and then evaluate in a systematic and robust scheme if these relevant DNA-probes are closely located, i.e. form a region of amplification/deletion.SIRAC does not need any preprocessing of the aCGH datasets, and requires only few, intuitive parameters.The results on two breast cancer datasets show promising outcomes that are in agreement with previous findings, but SIRAC better pinpoints the dissimilarities between the classes of interest.

View Article: PubMed Central - HTML - PubMed

Affiliation: Bioinformatics group, Delft University, Delft, The Netherlands. c.lai@tudelft.nl

ABSTRACT

Background: Array comparative genome hybridization (aCGH) provides information about genomic aberrations. Alterations in the DNA copy number may cause the cell to malfunction, leading to cancer. Therefore, the identification of DNA amplifications or deletions across tumors may reveal key genes involved in cancer and improve our understanding of the underlying biological processes associated with the disease.

Results: We propose a supervised algorithm for the analysis of aCGH data and the identification of regions of chromosomal alteration (SIRAC). We first determine the DNA-probes that are important to distinguish the classes of interest, and then evaluate in a systematic and robust scheme if these relevant DNA-probes are closely located, i.e. form a region of amplification/deletion. SIRAC does not need any preprocessing of the aCGH datasets, and requires only few, intuitive parameters.

Conclusion: We illustrate the features of the algorithm with the use of a simple artificial dataset. The results on two breast cancer datasets show promising outcomes that are in agreement with previous findings, but SIRAC better pinpoints the dissimilarities between the classes of interest.

Show MeSH
Related in: MedlinePlus