Limits...
Optimally discriminative subnetwork markers predict response to chemotherapy.

Dao P, Wang K, Collins C, Ester M, Lapuk A, Sahinalp SC - Bioinformatics (2011)

Bottom Line: However, prediction of response to cancer treatment has proved to be more challenging and novel approaches with improved generalizability are still highly needed.We show that our OptDis method improves over previously published subnetwork methods and provides better and more stable performance compared with other subnetwork and single gene methods.We also show that our subnetwork method produces predictive markers that are more reproducible across independent cohorts and offer valuable insight into biological processes underlying response to therapy.

View Article: PubMed Central - PubMed

Affiliation: School of Computing Science, Simon Fraser University.

ABSTRACT

Motivation: Molecular profiles of tumour samples have been widely and successfully used for classification problems. A number of algorithms have been proposed to predict classes of tumor samples based on expression profiles with relatively high performance. However, prediction of response to cancer treatment has proved to be more challenging and novel approaches with improved generalizability are still highly needed. Recent studies have clearly demonstrated the advantages of integrating protein-protein interaction (PPI) data with gene expression profiles for the development of subnetwork markers in classification problems.

Results: We describe a novel network-based classification algorithm (OptDis) using color coding technique to identify optimally discriminative subnetwork markers. Focusing on PPI networks, we apply our algorithm to drug response studies: we evaluate our algorithm using published cohorts of breast cancer patients treated with combination chemotherapy. We show that our OptDis method improves over previously published subnetwork methods and provides better and more stable performance compared with other subnetwork and single gene methods. We also show that our subnetwork method produces predictive markers that are more reproducible across independent cohorts and offer valuable insight into biological processes underlying response to therapy.

Availability: The implementation is available at: http://www.cs.sfu.ca/~pdao/personal/OptDis.html

Contact: cenk@cs.sfu.ca; alapuk@prostatecentre.com; ccollins@prostatecentre.com.

Show MeSH

Related in: MedlinePlus

Signaling pathways associated with TFAC response, ranked by enrichmentin the T50 SN derived from our OptDis method. We also compare the enrichment of those pathways in the genes from T50 SN (dark blue), O39 (light blue), T50 SG (cyan), and T111 SG markers (black). Significantly enriched pathways have Benjamini-Hochberg corrected p-values above threshold of 0.05 (dotted line).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3117373&req=5

Figure 6: Signaling pathways associated with TFAC response, ranked by enrichmentin the T50 SN derived from our OptDis method. We also compare the enrichment of those pathways in the genes from T50 SN (dark blue), O39 (light blue), T50 SG (cyan), and T111 SG markers (black). Significantly enriched pathways have Benjamini-Hochberg corrected p-values above threshold of 0.05 (dotted line).

Mentions: finally, we also compared subnetwork and single gene markers based on their insights into the mechanisms underlying drug response. We derived the T50 SN, T50 SG, and Tx SG from the combined cohort of 230 patients and used the Ingenuity Pathway Analysis software (IPA; Ingenuity© Systems, www.ingenuity.com) to identify significant pathway associations. Interestingly, several signaling pathways associated with chemotherapy response were identified for SN markers, whereas no significantly enriched pathways were found for the T50 and T111 SG markers (Fig. 6). A closer examination of the top associated pathways suggests response to TFAC treatment is affected by the cross-talk between tumor subtype specific mechanisms and pathways regulating apoptosis. Chemotherapy response in breast cancer have been observed to be subtype-specific (Sorlie, 2006), with ER+ tumors exhibiting much higher response rates to taxane-based therapies than ER− tumors (Farmer, 2009; Liedtke, 2008; Popovici, 2010). Therefore, it was expected to find that the predictive subnetwork signature was strongly enriched for genes activating the estrogen receptor (ER) signaling pathway. For the same reason, we also observe an enrichment for the androgen receptor (AR) signaling pathway. With nearly all ER+ tumors and few ER− tumors showing AR expression (Niemeier et al., 2010), it is likely that AR-based subnetworks serve as good predictive markers of TFAC treatment based on their association with ER status. Based on the enriched IPA pathways associated with response, we speculate that the differential response between subtypes may be attributed to differential regulation of apoptosis. Experimental studies have shown that expression of ERα selectively inhibits paclitaxel-induced apoptosis through modulation of glucocorticoid receptor activity (Sui et al., 2007).Fig. 6.


Optimally discriminative subnetwork markers predict response to chemotherapy.

Dao P, Wang K, Collins C, Ester M, Lapuk A, Sahinalp SC - Bioinformatics (2011)

Signaling pathways associated with TFAC response, ranked by enrichmentin the T50 SN derived from our OptDis method. We also compare the enrichment of those pathways in the genes from T50 SN (dark blue), O39 (light blue), T50 SG (cyan), and T111 SG markers (black). Significantly enriched pathways have Benjamini-Hochberg corrected p-values above threshold of 0.05 (dotted line).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3117373&req=5

Figure 6: Signaling pathways associated with TFAC response, ranked by enrichmentin the T50 SN derived from our OptDis method. We also compare the enrichment of those pathways in the genes from T50 SN (dark blue), O39 (light blue), T50 SG (cyan), and T111 SG markers (black). Significantly enriched pathways have Benjamini-Hochberg corrected p-values above threshold of 0.05 (dotted line).
Mentions: finally, we also compared subnetwork and single gene markers based on their insights into the mechanisms underlying drug response. We derived the T50 SN, T50 SG, and Tx SG from the combined cohort of 230 patients and used the Ingenuity Pathway Analysis software (IPA; Ingenuity© Systems, www.ingenuity.com) to identify significant pathway associations. Interestingly, several signaling pathways associated with chemotherapy response were identified for SN markers, whereas no significantly enriched pathways were found for the T50 and T111 SG markers (Fig. 6). A closer examination of the top associated pathways suggests response to TFAC treatment is affected by the cross-talk between tumor subtype specific mechanisms and pathways regulating apoptosis. Chemotherapy response in breast cancer have been observed to be subtype-specific (Sorlie, 2006), with ER+ tumors exhibiting much higher response rates to taxane-based therapies than ER− tumors (Farmer, 2009; Liedtke, 2008; Popovici, 2010). Therefore, it was expected to find that the predictive subnetwork signature was strongly enriched for genes activating the estrogen receptor (ER) signaling pathway. For the same reason, we also observe an enrichment for the androgen receptor (AR) signaling pathway. With nearly all ER+ tumors and few ER− tumors showing AR expression (Niemeier et al., 2010), it is likely that AR-based subnetworks serve as good predictive markers of TFAC treatment based on their association with ER status. Based on the enriched IPA pathways associated with response, we speculate that the differential response between subtypes may be attributed to differential regulation of apoptosis. Experimental studies have shown that expression of ERα selectively inhibits paclitaxel-induced apoptosis through modulation of glucocorticoid receptor activity (Sui et al., 2007).Fig. 6.

Bottom Line: However, prediction of response to cancer treatment has proved to be more challenging and novel approaches with improved generalizability are still highly needed.We show that our OptDis method improves over previously published subnetwork methods and provides better and more stable performance compared with other subnetwork and single gene methods.We also show that our subnetwork method produces predictive markers that are more reproducible across independent cohorts and offer valuable insight into biological processes underlying response to therapy.

View Article: PubMed Central - PubMed

Affiliation: School of Computing Science, Simon Fraser University.

ABSTRACT

Motivation: Molecular profiles of tumour samples have been widely and successfully used for classification problems. A number of algorithms have been proposed to predict classes of tumor samples based on expression profiles with relatively high performance. However, prediction of response to cancer treatment has proved to be more challenging and novel approaches with improved generalizability are still highly needed. Recent studies have clearly demonstrated the advantages of integrating protein-protein interaction (PPI) data with gene expression profiles for the development of subnetwork markers in classification problems.

Results: We describe a novel network-based classification algorithm (OptDis) using color coding technique to identify optimally discriminative subnetwork markers. Focusing on PPI networks, we apply our algorithm to drug response studies: we evaluate our algorithm using published cohorts of breast cancer patients treated with combination chemotherapy. We show that our OptDis method improves over previously published subnetwork methods and provides better and more stable performance compared with other subnetwork and single gene methods. We also show that our subnetwork method produces predictive markers that are more reproducible across independent cohorts and offer valuable insight into biological processes underlying response to therapy.

Availability: The implementation is available at: http://www.cs.sfu.ca/~pdao/personal/OptDis.html

Contact: cenk@cs.sfu.ca; alapuk@prostatecentre.com; ccollins@prostatecentre.com.

Show MeSH
Related in: MedlinePlus