Machine Learning Discoveries of TOP2A-X Synergy in ETC-1922159 Treated Colorectal Cancer Cells

1. Introduction

1.1. Topoisomerases

Becuase of the double-helical structure of the DNA, coils are often formed during relication, transcription and translation. This leads to building up of tension in the DNA structure. To maintain the relaxed topology of the DNA, topoisomerases act as essential proteins providing the required functionality. Wang [1] first discovered the existance of topoisomerases. In a review, McKie et al. [2] describe that topoisomerases are classified into different types depending on whether they catalyse the formation and re-ligation of single-stranded (ss) or double-stranded (ds)DNA breaks. A detailed description of the structure and function of DNA topoisomerases can be found in Champoux [3].

1.2. DNA Topoisomerase II

The ATP-dependent Type II which requires Mg²⁺, was identified in yeast by Goto and Wang [4]. Later, Adachi et al. [5] isolated the first conditional mutation in the mouse TOP2A. Watt and Hickson [6] particularly discuss the structure and function of type II. Of the two subtypes of type II, TOP2A has been found to be highly expressed in various malignancies, as reported by Zhou et al. [7]. In colon cancer, Zhang et al. [8] showed that the proliferation and invasion was suppressed by knockdown of TOP2A. Further, analysis by Carvalho et al. [9] revealed that TOP2A inhibitors were candidate drugs for rectal cancer treatment, based on drug repositioning. In colorectal cancer (CRC) cells treated with ETC-1922159, TOP2A was found to be down regulated along with other genes. TOP2A works in tandem with multiple components and some combinations of TOP2A have been confirmed in wet lab. However, many of the combinations have not been explored/tested or are known. To reveal these combinations, I use a modification of a recently published machine learning based search engine, details of which are given in the next section.

1.3. Combinatorial Search Problem and a Possible Solution

In a recently published work Sinha [10], a frame work of a search engine was developed which can rank combinations of factors (genes/proteins) in a signaling pathway. Readers are requested to go through the adaptation of the above mentioned work for gaining deeper insight into the working of the pipeline and its use of published data set generated after administration of ETC-1922159, Sinha [11]. The work uses SVM package by Joachims [12] in https://www.cs.cornell.edu/people/tj/svm_light/svm_rank.html. I use the adaptation to rank 2^nd order gene combinations.

2. Results & Discussion

2.1. TOP2A Related Synergies

2.1.1. TOP2A - WNT10B / NUSAP1 / WHSC1 / RHPN1-AS1 / ARID5B / TOPBP1

In non-small cell lung cancer, Wu et al. [13]’s experimental findings showed tha WNT3A, c-MYC, and $\beta$-catenin, expression levels were elevated when TOP2A was overexpressed and vice versa during TOP2A knockdown. Hu et al. [14] revealed that NUSAP1 gene silencing induced apoptosis in human glioblastoma, through the downregulation of the downstream molecule TOP2A. In hepatocellular carcinoma, Bao et al. [15] showed that activating transcription factor 2 (ATF2), bound to and promoted the transcription of WHSC1, which further increased the expression of TOP2A by inducing the dimethylation of histone H3 lysine 36 (H3K36me2). Zhou et al. [16] demostrated that RHPN1-AS1 negatively regulated miR-485-5p which lead to promotion of the TOP2A expression in ovarian cancer cells. Tsai et al. [17] show that ARID1A loss causes DNA replication stress associated with R-loops and transcription-replication conflicts in human cells. They show a model in which deletion of ARID1A reduces BRG1/BRM-associated factor (BAF) binding, thus failing to recruit TOP2A, which leads to accumulation of R-loops. Broderick et al. [18] identify TOPBP1 as an interactor of TOP2A, and show that it is required for TOP2A recruitment to resolve ultra-fine anaphase bridges (UFBs) during mitosis. These combinations with TOP2A which have been experimentally tested show a combinatorial synergy in various cases. In colorectal cancer cells treated with ETC-1922159, these components taken individually and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these individual members along with TOP2A.

Table 1 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 2 generated from analysis of the ranks in Table 1. The Table 1 shows rankings of individual members w.r.t TOP2A. WNT10B - TOP2A shows low ranking of 181 (laplace), 153 (linear) and 213 (rbf). NUSAP1 - TOP2A shows low ranking of 87 (laplace), 174 (linear) and 126 (rbf). WHSC1 - TOP2A shows low ranking of 444 (laplace), 711 (linear) and 541 (rbf). RHPN1-AS1 - TOP2A shows low ranking of 339 (laplace), 1381 (linear) and 371 (rbf). ARID5B - TOP2A shows low ranking of 464 (laplace), 855 (linear) and 355 (rbf). TOPBP1 - TOP2A shows low ranking of 702 (laplace), 1430 (linear) and 1032 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

One can also interpret the results of the Table 1 graphically, with the following influences - • individual members w.r.t TOP2A with TOP2A $->$ WNT10B / NUSAP1 / WHSC1 / RHPN1-AS1 / ARID5B / TOPBP1 .

2.1.2. TOP2A - ERCC

Polo-like kinase 1-interacting checkpoint helicase (PICH), also known as excision repair cross-complementation group 6 like (ERCC6L), is a substrate if PLK1-interacting checkpoint helicase. Overexpression of PICH/ERCC6L is related to the proliferation of tumors and Li et al. [19] hypothesis that PICH can maintain genomic stability by regulating appropriate chromosome structure, ensuring proper chromosome segregation, and facilitating replication fork reversal, via PICH-PLK1-TOP2A axis. In colorectal cancer cells treated with ETC-1922159, ERCC family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these ERCC member along with TOP2A.

Table 3 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 4 generated from analysis of the ranks in Table 3. The Table 3 shows rankings of individual members w.r.t TOP2A. ERCC6L - TOP2A shows low ranking of 40 (laplace), 24 (linear) and 24 (rbf). ERCC8 - TOP2A shows low ranking of 870 (laplace) and 903 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

One can also interpret the results of the Table 3 graphically, with the following influences - • ERCC family w.r.t TOP2A with TOP2A $->$ ERCC-6L/8.

2.1.3. TOP2A - EZH/CDK

In glioblastoma, Freitag et al. [20] show that EZH2 is overexpressed and combined EZH2-CDK4/6 inhibition increases antitumor activity against glioblastoma, by boosting cell death and cell stress, reverses stemness characteristics, disrupts endoplasmatic reticulum-mitochondrial homeostasis and reduces the invasion capability in GBM spheroids. Mechanistically, this was due to transcriptional changes in several genes, of which TOP2A is one of them. Slightly unrelated, but during retinal development in zebrafish Jin et al. [21] suggest that to regulate S phase entry, CDK1 interacts with cyclin A2 through phosphorylation of TOP2A. In colorectal cancer cells treated with ETC-1922159, EZH/CDK family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these EZH/CDK member along with TOP2A.

Table 5 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 6 generated from analysis of the ranks in Table 5. The Table 5 shows rankings of individual members w.r.t TOP2A. EZH2 - TOP2A shows low ranking of 1328 (laplace), 1463 (linear) and 1106 (rbf). CDK1 - TOP2A shows low ranking of 57 (laplace), 51 (linear) and 70 (rbf). CDK5RAP1 - TOP2A shows low ranking of 789 (laplace), 865 (linear) and 996 (rbf). CDK4 - TOP2A shows low ranking of 1277 (laplace), 948 (linear) and 1476 (rbf). CDK20 - TOP2A shows low ranking of 1370 (laplace), 1021 (linear) and 1421 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, CDK5RAP2 and CDK6 showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 5 graphically, with the following influences - • EZH-CDK family w.r.t TOP2A with TOP2A $->$ EZH2 and TOP2A $->$ CDK-1/5RAP1/4/20.

2.1.4. TOP2A - ORC

In primary and immortalized glioma cells, Yang et al. [22] showed that depleting/knockout of ORC6 decreased cell viability and proliferation, disrupted cell cycle progression and mobility, and triggered apoptosis. Further, via in vivo experiments, they demonstrated that ORC6 depletion decreased expression of Cyclin A2/B2/TOP2A. In colorectal cancer cells treated with ETC-1922159, ORC family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these ORC members along with TOP2A.

Table 7 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 8 generated from analysis of the ranks in Table 7. The Table 7 shows rankings of individual members w.r.t TOP2A. ORC1 - TOP2A shows low ranking of 258 (laplace), 145 (linear) and 279 (rbf). ORC6 - TOP2A shows low ranking of 1296 (laplace), 490 (linear) and 1449 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, ORC5, ORC3 and ORC2 showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 7 graphically, with the following influences - • ORC family w.r.t TOP2A with TOP2A $->$ ORC-1/6.

2.1.5. TOP2A - IL

Through in vivo and in vitro experiments Li et al. [23] showed that knockdown of TOP2A inhibited inflammation and IL-17 signaling pathway, and promoted proliferation of ulcerative colitis (an inflammatory disease of the colonic mucosa). In colorectal cancer cells treated with ETC-1922159, IL family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these IL members along with TOP2A.

Table 9 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 10 generated from analysis of the ranks in Table 9. The Table 9 shows rankings of individual members w.r.t TOP2A. IL33 - TOP2A shows low ranking of 591 (laplace), 152 (linear) and 818 (rbf). IL17D - TOP2A shows low ranking of 1054 (laplace), 1194 (linear) and 853 (rbf). IL17RD - TOP2A shows low ranking of 1193 (laplace) and 1260 (rbf). IL1RL2 - TOP2A shows low ranking of 1237 (laplace) and 1328 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

One can also interpret the results of the Table 9 graphically, with the following influences - • IL family w.r.t TOP2A with TOP2A $->$ IL-33/17D/17RD/1RL2.

2.1.6. TOP2A - USP

Fielding et al. [24] show that deubiquitylase USP15 is required for TOP2A accumulation during G2, and USP15 depletion causes formation of anaphase chromosome bridges. In colorectal cancer cells treated with ETC-1922159, USP family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these USP members along with TOP2A.

Table 11 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 12 generated from analysis of the ranks in Table 11. The Table 11 shows rankings of individual members w.r.t TOP2A. USP13 - TOP2A shows low ranking of 49 (laplace), 63 (linear) and 43 (rbf). USP36 - TOP2A shows low ranking of 964 (laplace) and 1077 (rbf). USP28 - TOP2A shows low ranking of 1147 (laplace), 1546 (linear) and 1004 (rbf). USP1 - TOP2A shows low ranking of 1375 (linear) and 1241 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, USP39 and USP10 showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 11 graphically, with the following influences - • USP family w.r.t TOP2A with TOP2A $->$ USP-13/36/28/1.

2.1.7. TOP2A - RAD54-ZNF

TOP2 form cleavage complexes (TOP2ccs) during their catalytic cycle to relieve topological stress, however they can be trapped by TOP2 poisons. Trapped TOP2ccs by action of TOP2 poisons, block transactions on DNA and generate genotoxic stress. Zhang et al. [25] uncovered RAD54L2 which mediates a TOP2-specific DNA damage avoidance pathway, by interacting with TOP2A/TOP2B and ZATT/ZNF451. In colorectal cancer, Gao et al. [26] showed that ZNF148 modulates TOP2A. In colorectal cancer cells treated with ETC-1922159, RAD54-ZNF family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these RAD54-ZNF members along with TOP2A.

Table 13 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 14 generated from analysis of the ranks in Table 13. The Table 13 shows rankings of individual members w.r.t TOP2A. RAD54B - TOP2A shows low ranking of 145 (laplace), 188 (linear) and 80 (rbf). RAD54L - TOP2A shows low ranking of 1233 (laplace) and 933 (rbf). ZNF815P - TOP2A shows low ranking of 45 (laplace), 72 (linear) and 38 (rbf). ZNF204P - TOP2A shows low ranking of 74 (laplace), 165 (linear) and 139 (rbf). ZNF367 - TOP2A shows low ranking of 106 (laplace), 37 (linear) and 96 (rbf). ZNF775 - TOP2A shows low ranking of 194 (laplace), 110 (linear) and 204 (rbf). ZNF485 - TOP2A shows low ranking of 309 (laplace), 936 (linear) and 349 (rbf). ZNF239 - TOP2A shows low ranking of 556 (laplace), 956 (linear) and 411 (rbf). ZNF511 - TOP2A shows low ranking of 621 (laplace), 612 (linear) and 737 (rbf). ZNF519 - TOP2A shows low ranking of 629 (laplace), 264 (linear) and 536 (rbf). ZNF771 - TOP2A shows low ranking of 677 (laplace), 636 (linear) and 542 (rbf). ZNF74 - TOP2A shows low ranking of 714 (laplace), 1371 (linear) and 726 (rbf). ZNF202 - TOP2A shows low ranking of 775 (laplace), 1331 (linear) and 782 (rbf). ZNF620 - TOP2A shows low ranking of 784 (laplace), 600 (linear) and 620 (rbf). ZNF22 - TOP2A shows low ranking of 826 (laplace), 666 (linear) and 735 (rbf). ZNF273 - TOP2A shows low ranking of 866 (laplace), 1684 and 935 (rbf). ZNF584 - TOP2A shows low ranking of 926 (laplace), 469 (linear) and 1120 (rbf). ZNF248 - TOP2A shows low ranking of 957 (laplace) and 1085 (rbf). ZNF90 - TOP2A shows low ranking of 975 (laplace), 968 (linear) and 1123 (rbf). ZNF740 - TOP2A shows low ranking of 1061 (laplace), 1219 (linear) and 1485 (rbf). ZNF138 - TOP2A shows low ranking of 1092 (laplace) and 1277 (rbf). ZNF48 - TOP2A shows low ranking of 1207 (laplace), 653 (linear) and 1477 (rbf). ZNF512 - TOP2A shows low ranking of 1214 (laplace), 807 (linear) and 1375 (rbf). ZNF695 - TOP2A shows low ranking of 1228 (laplace), 858 (linear) and 1393 (rbf). ZNF793 - TOP2A shows low ranking of 1243 (laplace), 1073 (linear) and 1139 (rbf). ZNF572 - TOP2A shows low ranking of 1257 (laplace), 1503 (linear) and 1481 (rbf). ZNF670 - TOP2A shows low ranking of 1258 (laplace), 541 (linear) and 1115 (rbf). ZNF32 - TOP2A shows low ranking of 1332 (laplace), 724 (linear) and 1213 (rbf). ZNF232 - TOP2A shows low ranking of 1344 (laplace) and 1098 (rbf). ZNF691 - TOP2A shows low ranking of 1394 (laplace), 1271 (linear) and 1172 (rbf). ZNF124 - TOP2A shows low ranking of 1361 (linear) and 1351 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, ZNF37BP, ZNF486, ZNF594, ZNF146, ZNF589, ZNF280D, ZNF738, ZNF736, ZNF428, ZNF330, ZNF174, ZNF780B, ZNF93, ZNF629, ZNF614, ZNF346, ZNF639, ZNF263, ZNF215, ZNF275, ZNF343, ZNF480, ZNF789, ZNF502, ZNF445 and ZNF689 showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 13 graphically, with the following influences - • RAD54-ZNF family w.r.t TOP2A with TOP2A $->$ RAD54-B/L and TOP2A $->$ ZNF-815P / 204P / 367 / 775 / 485 / 239 / 511 / 519 / 771 / 74 / 202 / 620 / 22 / 273 / 584 / 248 / 90 / 740 / 138 / 48 / 512 / 695 / 793 / 572 / 670 / 32 / 232 / 691 / 124.

2.1.8. TOP2A - HMGB

In carbon tetrachloride (CCl₄) induced liver fibrosis mouse model, Huang et al. [27] showed that deletion/inhibition of HMGB2 slowed the progression of CCl₄- induced liver fibrosis. Their RNA-seq analysis revealed the induction of CCl₄-activated genes, one of which was TOP2A (the activation of which was abolished in HMGB2^−/− mice or in ICM-treated mice). In colorectal cancer cells treated with ETC-1922159, HMGB family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these HMGB members along with TOP2A.

Table 15 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 16 generated from analysis of the ranks in Table 15. The Table 15 shows rankings of individual members w.r.t TOP2A. HMGB2 - TOP2A shows low ranking of 378 (laplace), 330 (linear) and 314 (rbf). HMGB3 - TOP2A shows low ranking of 987 (laplace), 1173 (linear) and 827 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, HMGB1 showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 15 graphically, with the following influences - • HMGB family w.r.t TOP2A with TOP2A $->$ HMGB-2/3.

2.1.9. TOP2A - E2F

In gastric cancer cells Chen et al. [28] showed that overexpressed E2F1 increased TOP2A levels and vice versa. In colorectal cancer cells treated with ETC-1922159, E2F family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these E2F members along with TOP2A.

Table 17 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 18 generated from analysis of the ranks in Table 17. The Table 17 shows rankings of individual members w.r.t TOP2A. E2F7 - TOP2A shows low ranking of 7 (laplace), 20 (linear) and 8 (rbf). E2F8 - TOP2A shows low ranking of 291 (laplace), 85 (linear) and 373 (rbf). E2F2 - TOP2A shows low ranking of 436 (laplace), 219 (linear) and 550 (rbf). E2F1 - TOP2A shows low ranking of 606 (laplace), 284 (linear) and 562 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, E2F5 showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 17 graphically, with the following influences - • E2F family w.r.t TOP2A with TOP2A $->$ E2F-7/8/2/1.

2.1.10. TOP2A - GINS-MCM

GINS1 is a GINS complex subunit that interacts with the MCM2-7 complex and CCD45 in eukaryotic DNA replication. Gambus et al. [29] show that the GINS complex allows MCM helicase to interact with replisome progression complexes (RPCs) that are assembled during initiation and disassembled at the end of S phase. RPCs also interact with MCM10 and TOP1. In glioma cells and tissues, Yang et al. [30] showed that GINS1 expression level was upregulated and mechanistically it promotes proliferation and migration through USP15-mediated deubiquitination of TOP2A protein. In colorectal cancer cells treated with ETC-1922159, GINS-MCM family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these GINS-MCM members along with TOP2A.

Table 19 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 20 generated from analysis of the ranks in Table 19. The Table 19 shows rankings of individual members w.r.t TOP2A. For GINS - GINS1 - TOP2A shows low ranking of 102 (laplace), 78 (linear) and 173 (rbf). GINS2 - TOP2A shows low ranking of 131 (laplace), 49 (linear) and 140 (rbf). GINS3 - TOP2A shows low ranking of 411 (laplace), 312 (linear) and 417 (rbf). GINS4 - TOP2A shows low ranking of 601 (laplace), 406 (linear) and 375 (rbf).

For MCM - MCM4 - TOP2A shows low ranking of 52 (laplace), 66 (linear) and 49 (rbf). MCM10 - TOP2A shows low ranking of 113 (laplace), 301 (linear) and 264 (rbf). MCM2 - TOP2A shows low ranking of 236 (laplace), 384 (linear) and 237 (rbf). MCM5 - TOP2A shows low ranking of 256 (laplace), 590 (linear) and 224 (rbf). MCM6 - TOP2A shows low ranking of 467 (laplace), 359 (linear) and 386 (rbf). MCM8 - TOP2A shows low ranking of 474 (laplace), 458 (linear) and 592 (rbf). MCM3 - TOP2A shows low ranking of 487 (laplace), 794 (linear) and 268 (rbf). MCM7 - TOP2A shows low ranking of 533 (laplace), 801 (linear) and 438 (rbf). These rankings point to the synergy existing between the two/three components, which have been down regulated after the drug treatment.

One can also interpret the results of the Table 19 graphically, with the following influences - • GINS-MCM family w.r.t TOP2A with TOP2A $->$ GINS-1/2/3/4 and TOP2A $->$ MCM-4/10/2/5/6/8/3/7.

2.1.11. TOP2A - BUB

Carvalhal et al. [31] confirm that BUB1 kinase activity promotes the centromeric localization of TOP2A. In colorectal cancer cells treated with ETC-1922159, BUB family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these BUB members along with TOP2A.

Table 21 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 22 generated from analysis of the ranks in Table 21. The Table 21 shows rankings of individual members w.r.t TOP2A. BUB1 - TOP2A shows low ranking of 19 (laplace), 41 (linear) and 44 (rbf). BUB1B - TOP2A shows low ranking of 211 (laplace), 154 (linear) and 99 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, BUB3 showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 21 graphically, with the following influences - • BUB family w.r.t TOP2A with TOP2A $->$ BUB-1/1B.

2.1.12. TOP2A - DDX

Zhang et al. [32] suggest that DDX11-AS1 knockdown resulted in reduced resistance of esophageal cancer cells to paclitaxel by inhibiting TOP2A transcription via TAF1. In colorectal cancer cells treated with ETC-1922159, DDX family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these DDX members along with TOP2A.

Table 23 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 24 generated from analysis of the ranks in Table 23. The Table 23 shows rankings of individual members w.r.t TOP2A. DDX11-AS1 - TOP2A shows low ranking of 310 (laplace), 365 (linear) and 406 (rbf). DDX12P - TOP2A shows low ranking of 435 (laplace), 76 (linear) and 450 (rbf). DDX55 - TOP2A shows low ranking of 806 (laplace), 1389 (linear) and 919 (rbf). DDX11 - TOP2A shows low ranking of 829 (laplace) and 813 (rbf). DDX28 - TOP2A shows low ranking of 965 (laplace), 1108 (linear) and 871 (rbf). DDX18 - TOP2A shows low ranking of 1084 (laplace) and 1273 (rbf). DDX20 - TOP2A shows low ranking of 1369 (laplace), 393 (linear) and 1401 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, DDX54, DDX10, DDX31, DDX21, DDX51, DDX19A, DDX56, DDX46 and DDX27 showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 23 graphically, with the following influences - • DDX family w.r.t TOP2A with TOP2A $->$ DDX-11-AS1/12P/55/11/28/18/20.

2.1.13. TOP2A - H2A

Zhang et al. [33] show that histone H2A phosphorylation modification generated by the mitotic kinase BUB1, is necessary and sufficient for the centromeric localization of TOP2A. In colorectal cancer cells treated with ETC-1922159, H2A family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these H2A members along with TOP2A.

Table 25 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 26 generated from analysis of the ranks in Table 25. The Table 25 shows rankings of individual members w.r.t TOP2A. H2AFV - TOP2A shows low ranking of 409 (laplace) and 798 (rbf). H2AFZ - TOP2A shows low ranking of 1178 (laplace) and 1108 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, H2AFX showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 25 graphically, with the following influences - • H2A family w.r.t TOP2A with TOP2A $->$ H2A-FV/FZ.

2.1.14. TOP2A - SMC

The TOP2A-dependent arrest is responsible ofr segregation of sister chromatids and has been identified as dysfunctional in various tumour cell lines. Deiss et al. [34] show that the SMC5/6 complex regulates the TOP2A-dependent G2 arrest and sister chromatid disjunction via NSE2-mediated SUMOylation of TOP2A. In colorectal cancer cells treated with ETC-1922159, SMC family members and TOP2A, were found to be down regulated and their regulation was recorded independently. I was able to rank 2^nd order combination of these SMC members along with TOP2A.

Table 27 shows rankings of these combinations. Followed by this is the unexplored combinatorial hypotheses in Table 28 generated from analysis of the ranks in Table 27. The Table 27 shows rankings of individual members w.r.t TOP2A. SMC2 - TOP2A shows low ranking of 335(laplace), 970 (linear) and 340 (rbf). SMC4 - TOP2A shows low ranking of 915(laplace), 446 (linear) and 986 (rbf). These rankings point to the synergy existing between the two components, which have been down regulated after the drug treatment.

Further, SMC1A showed high ranking with TOP2A, thus indicating that they might not be working synergistically with TOP2A, before the drug treatment.

One can also interpret the results of the Table 27 graphically, with the following influences - • SMC family w.r.t TOP2A with TOP2A $->$ SMC-2/4.

3. Conclusion

Presented here are a range of multiple synergistic TOP2A 2^nd order combinations that were ranked via a machine learning based search engine. Via majority voting across the ranking methods, it was possible to find plausible unexplored synergistic combinations of TOP2A-X that might be prevalent in CRC cells after treatment with ETC-1922159 drug.