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

1. Introduction

In the unpublished preprint [1], a frame work of a search engine was developed which can rank combinations of factors (genes/proteins) in a signaling pathway. Such combinations are of import due to the vast search space in which they exist and the difficulty to find them. The search engine facilitates in prioritizing the combinations as ranked biological hypotheses which the biologists might want to test in wet lab, to know if a synergistic combination is prevalent in a signaling pathway, in a direct or indirect manner. Interested readers are advised to go through unpublished preprints [1,2] for details regarding the search engine and the discoveries mentioned in there.

2. Materials and Methods

2.1. Combinatorial Search Problem and a Possible Solution

The issue of combinatorial search problem and a possible solution has been addressed in [2,3]. The details of the methodology of this manuscript have been explained in great detail in [3] & its application in [2]. Readers are requested to go through the same for gaining deeper insight into the working of the pipeline and its use of published data set generated after administration of ETC-1922159. In order to understand the significance of the solution proposed to the problem of combinatorial search that the biologists face in revealing unknown biological search problem, these works are of importance.

Briefly, from [2], the pipleline works by computing sensitivity indicies for each of these unique combinations and then vectorising these indices to connote and form discriminative feature vector for each combination. Since each combination is unique, the training and the test data are same. In the training data, the combinations are arranged and ranks from 1 to n are assigned. The ranking algorithm then learns the patterns from these combinations/sensitivity index vectors. Next the learned model is used to rank the test data by generating the ranking score for each of the unique combination. Sorting these shuffled scores of test data leads to prioritization of the combinations. [4] show an example of applying learned model to training data (same as the test data) in https://www.cs.cornell.edu/people/tj/svm_light/svm_rank.html. Note that these combinations are now ranked and give the biologists a chance to narrow down their focus on crucial biological hypotheses in the form of combinations which the biologists might want to test. Analogous to the webpage search engine, where the click of a button for a few key-words leads to a ranked list of web links, the pipeline uses sensitivity indices as an indicator of the strength of the influence of factors or their combinations, as a criteria to rank the combinations.

3. Results & Discussion

3.1. WNT Related Synergies

3.1.1. WNT10B-ASCL2

WNT10B has been found to be implicated in a range of cancers. In gastric cancer, the knockdown of WNT10B showed reduced expression of cell proliferation and migration as well as inhibition of epithelial-mesenchymal transition [5]. On the other hand, WNT10B is also involved in the formation of bone mass and progenitor maintenance of various kinds of tissue, while deletion of the same leads to loss of bone mass and mesenchymal progenitor cells [6]. Their contribution is also reported in axonal regeneration in injured CNS [7]. Furthermore, like WNT10B, WNT10A and WNT6 have shown to play a major role in inhibiting adipogenesis and stimulates osteoblastogenesis while regulating the mesenchymal stem cells [8,9]. Involvement in heptocellular carcinoma of WNT10B has been found wherein it is shown that stable silencing of WNT10B leads to significant reduction in proliferation, colony formation, migration and invasion in HepG2 HCC cell line [10]. Its implication in breast cancer [11,12] as well as endometrial cancer [13] has also been reported.

In colorectal cancer, WNT10B has shown to play a dual function of both oncogenesis promotion via $\beta$-catenin/TCF pathway and the inhibition of cell growth, possibly via FGF family of proteins [14]. Methylation of WNT10B has been found in the some of the cancer cell lines while its reversal has lead to over-expression of the WNT10B. However, the over-expression of WNT10B has lead to reduced cell growth in cancer, indicating a $\beta$-catenin independent component to be behind such a phenomena. Methylation of over-expressed WNT10B and synergistic work with FGF family of proteins later indicate the promotion of oncogenesis, as has been demonstrated in [14].

In a more recent work, ASCL2 has been found to play a major role in stemness in colon crypts and is implicated in colon cancer [15]. Switching off the ASCL2 leads to a literal blockage of the stemness process and vice versa. At the downstream level, ASCL2 is regulated by TCF4/$\beta$-catenin via non-coding RNA target named WiNTRLINC1 [16]. Activation of ASCL2 leads to feedforward transcription of the non-coding RNA and thus a loop is formed which helps in the stemness and is highly effective in colon cancer. At the upstream level, ASCL2 is known act as a WNT/RSPONDIN switch that controls the stemness [17]. It has been shown that removal of RSPO1 lead to decrease in the Wnt signaling due to removal of the FZD receptors that led to reduced expression of ASCL2. Also, low levels of LGR5 were observed due to this phenomena. The opposite happened by increasing the RSPO1 levels. After the drug treatment, it was found that ASCL2 was highly suppressed pointing to the inhibition of stemness in the colorectal cancer cells. Also, [17] show that by genetically disrupting PORCN or inducing a PORCN inhibitor (like IWP-2), there is loss of stem cell markers like LGR5 and RNF43, which lead to disappearance of stem cells and moribund state of mice. A similar affect can be found with ETC-1922159, where there is suppression of RNF43 and LGR5 that lead to inhibition of the Wnt pathway and thus the ASCL2 regulation. These wet lab evidences are confirmed in the relatively low ranking of the combination ASCL2-RNF43 via the inhibition of PORCN-WNT that leads to blocking of the stemness that is induced by ASCL2. Since ASCL2 is directly mediated by the WNT proteins, the recorded ASCL2-WNT10B combination showed low priority ranking of 488, 497 and 321 for rbf, laplace and linear kernels, respectively, thus indicating a possible connection between WNT10B and ASCL2 activation. WNT10B might be playing a crucial role in stemness. This is further confirmed by wet lab experiments in [18], which show BVES deletion results in amplified stem cell activity and Wnt signaling after radiation. WNT10B has been implicated in colorectal cancer [14].

3.1.2. ABC Transporters - WNT Cross Family Analysis

[19] have shown the role of ABC transporters in progression and clinical outcome of colorectal cancer. Work by [20] show that Wnt-$\beta$ catenin signaling regulates ABCC3 (MRP3) transporter expression in colorectal cancer. ABCA2 belongs to the category of ABC transporters that play an essential role in the development of resistance by the efflux of anticancer agents outside of cancer cells [19]. [19] observed that ABCA2 had no significant change/affect in colorectal cancer cases. [20] found ABCA2 to be downregulated in colorectal cancer case. In ETC-1922159 affected CRC cells, down regulation of ABCA2 was observed, after the inhibition of proliferation in respective cells. Multiple members of ABC transporters and WNTs were found to be UP regulated after ETC-159 in CRC cells and WNTs are known to regulate ABCs. Below, we show a range of up regulated, possible unknown and unexplored synergistic 2^nd order combinations that were ranked by the search engine. Note that the high numerical valued ranks (i.e nearing to 1800/2000 and above) indicate high potential of synergy that might be existing in CRC cells after the drug administration. Majority voting of rankings across the three different kernels point to the potential of the synergistic discovery. Wet labs investigations will assist in confirmation of these discoveries and if proven true, might lead to understanding of further mechanism between the components.

Table 1 and Table 2 show the rankings of ABC family w.r.t to WNT family members and WNT family w.r.t to ABC family members, respectively. From these two tables, we derive the plausible influences that might be existing in a two way format that is depicted in Table 3. In Table 1, WNT2B - ABC-C3 combination shows a majority voting of 1853 (laplace) and 2498 (rbf). Similarly, WNT7B - ABC-C13 shows a majority voting of 2245 (linear) and 2298 (rbf). These two combinations are depicted in Table 3 as ABC members influenced by WNT members (see under ABC w.r.t WNT). Reversibily, in Table 2 ABC-A5 - WNT2B shows a majority voting of 2018 (linear) and 2132 (rbf), ABC-A5 - WNT4 shows a majority voting of 2436 (linear) and 2449 (rbf), ABC-A5 - WNT9A shows a majority voting of 1989 (laplace), 2209 (linear) and 2365 (rbf), WNT2B - ABC-C5 shows a majority voting of 1970 (laplace), 2309 (linear) and 2248 (rbf), ABC-C5 - WNT9A shows a majority voting of 2183 (linear) and 2480 (rbf), WNT2B - ABC-C13 shows a majority voting of 2150 (linear) and 2048 (rbf), WNT7B - ABC-C13 shows a majority voting of 2508 (laplace) and 1830 (linear), WNT7B - ABC-D1 shows a majority voting of 2238 (laplace) and 2021 (linear), WNT7B - ABC-G1 shows a majority voting of 1808 (linear) and 1866 (rbf), WNT7B - ABC-G2 shows a majority voting of 2334 (linear) and 2145 (rbf) and WNT9A - ABC-G2 shows a majority voting of 1919 (laplace) and 2003 (rbf). These point to WNT members influenced by ABC members (see under WNT w.r.t ABC). Hypothetically, what we find is that the synergies can be bi-directional also and might contain various intermitent factors through which the factors might be working synergistically. These hypothese form present themselves as important combinations that might be of interest to biologists/oncologists.

One can also interpret the results of the Table 3 graphically, with the following influences - • ABC w.r.t WNT with WNT-2B $->$ ABC-C3; WNT-7B $->$ ABC-C13; and • WNT w.r.t ABC with ABC-A5 $<-$ WNT-2B/4/9A; WNT-2B/9A $<-$ ABC-C5; WNT-2B/7B $<-$ ABC-C13; WNT-7B $<-$ ABC-D1; WNT-7B $<-$ ABC-G1; WNT-7B/9A $<-$ ABC-G2. Thus, in this way, we can utilize the search engine to derive the various probable combinations between the factors of interest and their interdependent influences through the two-way cross family analysis.

3.1.3. IL - WNT Cross Family Analysis

Interleukin (IL) has been found in cross talk with WNT pathway. [21] show that NF$\kappa$B induced WNT signaling in colorectal cancer via interleukin-1$\beta$ IL1B. Further, [22] have shown that nitric oxide mediates crosstalk between interleukin 1$\beta$ and Wnt signaling in primary human chondrocytes by reducing DKK1 and FRZB expression. The role of IL-17 (Interleukin-17) family is known to be controversial in CRC, however there are cases were it has been reported to be a prognostic marker for colorectal cancer [23,24]. A homologue of the family, IL-17D a novel cytokine has been discovered [25] and found to play a role in many of the cancers. In cells treated with ETC-1922159, IL-17D was found to be down regulated and reversibly it must have been regulated in the colorectal cancer cases. Recently, crosstalk between WNT/$\beta$-Catenin and NF-$\kappa$B signaling pathway during inflammation has been reported by [26]. [27] also show WNT/$\beta$-catenin negative feedback loop inhibits IL-1 induced matrix metalloproteinase expression in human articular chondrocytes. [28] conclude that WNT/$\beta$-catenin signaling promotes angiogenesis possibly via the induction of known angiogenic regulators such as Interleukin-8. In mouse colon, Interleukin-1 signaling is shown to mediate obesity-promoted elevations in inflammatory cytokines, WNT activation, and epithelial proliferation by [29]. In pulmonary fibrosis, [30] show that WNT/$\beta$-Catenin signaling induces IL-1$\beta$ expression by alveolar epithelial cells. [31] show that IL-23 promotes the epithelial-mesenchymal transition of oesophageal carcinoma cells via the WNT/$\beta$-catenin pathway. Finally, [32] show that IL-6/WNT interactions in rheumatoid arthritis.

Family members belonging to each of the factors like WNT, IL etc, might be involved synergistically in pathological case or otherwise. IL and WNT members were found to be up regulated after the treatment of ETC-1922159 in colorectal cancer cells. We present here, multiple plausible and alternative synergistic combinatorial biological hypotheses for IL-WNT combination, which emerge after a cross family member analysis of the in silico revelations pertaining to the components under investigation.

Table 4 shows IL-WNT two way cross family analysis. The left side of the table contains rankings of IL family with respect to WNTs and the right side of the table contains rankings of WNT family with respect to ILs. Depicted in table are the plausible combinatorial hypotheses derived from majority voting of the rankings in Table 4. On the left half, w.r.t WNT2B, IL-6ST/8/17REL show a synergy with WNT2B. These are reflected with rankings of 1797 (linear) and 2088 (rbf) for IL-6ST - WNT2B; rankings of 2107 (laplace), 1817 (linear) and 2088 (rbf) for IL-8 - WNT2B and rankings of 1824 (laplace) and 2241 (rbf) for IL-17REL - WNT2B, respectively. W.r.t WNT4, IL-1B/1RAP/15RA/17C show a synergy with WNT4. These are reflected with rankings of 1867 (laplace) and 1976 (linear) for IL-1B - WNT4; rankings of 2302 (laplace) and 1826 (linear) for IL-1RAP - WNT4; rankings of 1987 (laplace) and 2265 (linear) for IL-15RA - WNT4 and rankings of 2018 (laplace) and 1881 (linear) for IL-17C - WNT4, respectively. W.r.t WNT7B, IL-1RN/17REL show a synergy with WNT7B. These are reflected with rankings of 1882 (laplace) and 1796 (linear) for IL-1RN - WNT7B and rankings of 2053 (laplace), 2445 (linear) and 2489 (rbf) for IL-17REL - WNT4, respectively. W.r.t WNT9A, IL-1RAP/15RA show a synergy with WNT9A. These are reflected with rankings of 2273 (linear) and 2159 (rbf) for IL-1RAP - WNT9A and rankings of 1776 (laplace) and 2380 (linear) for IL-15RA - WNT9A, respectively.

On the right half, WNT2B w.r.t IL family, IL-1A/1RAP/8 show a synergy with WNT2B. These are reflected with rankings of 2290 (laplace) and 2427 (rbf) for IL-1A - WNT2B; rankings of 2488 (laplace) and 1892 (rbf) for IL-1RAP - WNT2B and rankings of 2157 1824 (laplace) and 2025 (linear) for IL-8 - WNT2B, respectively. WNT4 w.r.t IL family, IL-8/10RB show a synergy with WNT4. These are reflected with rankings of 1980 (laplace) and 2144 (linear) for IL-8 - WNT4 and rankings of 1828 (laplace), 2259 (linear) and 1993 (rbf) for IL-10RB - WNT4; respectively. WNT7B w.r.t IL family, IL-1A/1RN/6ST/17C show a synergy with WNT7B. These are reflected with rankings of 2134 (linear) and 2312 (rbf) for IL-1A - WNT7B; rankings of 1907 (laplace) and 2162 (linear) for IL-1RN - WNT7B; rankings of 1881 (linear) and 2020 (rbf) for IL-ST - WNT7B; and rankings of 1956 (laplace), 2388 (linear) and 1982 (rbf) for IL-17C - WNT7B, respectively. WNT9A w.r.t IL family, IL-1RAP/15RA/17REL show a synergy with WNT9A. These are reflected with rankings of 2003 (laplace) and 2179 (linear) for IL-1RAP - WNT9A; rankings of 2149 (laplace) and 2362 (linear) for IL-15RA - WNT9A; and rankings of 2101 (laplace) and 1940 (linear) for IL-17REL - WNT9A, respectively. One can also interpret the results of the Table 5 graphically, with the following influences - • IL w.r.t WNT with IL-6ST/8/17REL $<-$ WNT-2B; IL-1B/1RAP/15RA/17C $<-$ WNT-4; IL-1RN/17REL $<-$ WNT-7B; IL-1RAP/15RA $<-$ WNT-9A and • WNT w.r.t IL with IL-1A/1RAP/8 $->$ WNT-2B; IL-8/10RB $->$ WNT-4; IL-1A/1RN/6ST/17C $->$ WNT-7B and IL-1RAP/15RA/17REL $->$ WNT-9A.

3.1.4. UBE2 - WNT Cross Family Analysis

[33] observed balanced ubiquitylation and deubiquitylation of Frizzled regulate cellular responsiveness to Wg/Wnt. Family members belonging to each of the factors like UBE2, WNT etc, might be involved synergistically in pathological case or otherwise. UBE2 and WNT members were found to be up regulated after the treatment of ETC-159 in colorectal cancer cells. However, not much is known about interation between the UBE2 family members and WNTs. Here we present a range of synergies that were ranked highly for up regulation. Table 6 presents the rankings of UBE family VS WNT family. Following this, is the Table 7 which derives the necessary influences via majority voting of rankings in Table 6.

On the left half, w.r.t WNT family, UBE2A show a synergy with WNT4. These are reflected with rankings of 2314 (linear) and 2279 (rbf) for UBE2A - WNT4; UBE2B show a synergy with WNT4/7B. These are reflected with rankings of 2260 (laplace), 2008 (linear) and 2141 (rbf) for UBE2B - WNT4 and rankings of 2116 (laplace) and 2206 (rbf) for UBE2B - WNT7B, respectively; UBE2F show a synergy with WNT4/7B. These are reflected with rankings of 2135 (laplace) and 2505 (linear) for UBE2F - WNT4 and rankings of 2423 (laplace) and 2077 (rbf) for UBE2F - WNT7B, respectively; UBE2H show a synergy with WNT2B. These are reflected with rankings of 1841 (laplace) and 2178 (linear) for UBE2H - WNT2B; UBE2J1 show a synergy with WNT-7B/9A. These are reflected with rankings of 2349 (laplace) and 2183 (rbf) for UBE2J1 - WNT7B and rankings of 1835 (laplace) and 2053 (rbf) for UBE2J1 - WNT9A, respectively. UBE2Z show a synergy with WNT-2B/4/9A. These are reflected with rankings of 1756 (linear) and 1878 (rbf) for UBE2J1 - WNT2B, rankings of 2195 (laplace) and 2468 (rbf) for UBE2J1 - WNT4, and 2343 (laplace) and 1973 (rbf) for UBE2J1 - WNT9A, respectively.

On the right half, w.r.t UBE2, UBE2A shows a synergy with WNT4. These are reflected with rankings of 2345 (linear) and 2151 (rbf) for UBE2A - WNT7B; UBE2B shows a synergy with WNT-7B/9A. These are reflected with rankings of 2052 (linear) and 1903 (rbf) for UBE2B - WNT7B and rankings of 2300 (laplace), 2476 (linear) and 2326 (rbf) for UBE2B - WNT9A, respectively; UBE2F shows a synergy with WNT-7B/9A. These are reflected with rankings of 2236 (laplace) and 1751 (rbf) for UBE2F - WNT7B and rankings of 2251 (linear) and 2179 (rbf) for UBE2F - WNT9A, respectively; UBE2H shows a synergy with WNT4. These are reflected with rankings of 2248 (linear) and 2155 (rbf) for UBE2H - WNT4; UBE2J1 shows a synergy with WNT-7B/9A. These are reflected with rankings of 1877 (llinear) and 1846 (rbf) for UBE2J1 - WNT7B and rankings of 2471 (laplace), 2137 (linear) and 2469 (rbf) for UBE2J1 - WNT9A, respectively. UBE2Z shows a synergy with WNT-9A. These are reflected with rankings of 1972 (laplace) and 1800 (linear) for UBE2Z - WNT7B, respectively.

One can also interpret the results of the Table 7 graphically, with the following influences - • UBE2 w.r.t WNT with WNT-4 $->$ UBE2-A; WNT-4/7 $->$ UBE2-B; WNT-4/7B $->$ UBE2-F; WNT-2B $->$ UBE2-H; WNT-7B/9B $->$ UBE2-J1; WNT-2B/4/7B $->$ UBE2-Z and • WNT w.r.t UBE2 with WNT-7B $<-$ UBE2-A; WNT-7B/9A $<-$ UBE2-B; WNT-7B/9A $<-$ UBE2-F; WNT-4 $<-$ UBE2-H; WNT-7B/9A $<-$ UBE2-J1; WNT-7B $<-$ UBE2-Z;

3.1.5. EXOSC - WNT10B Cross Family Analysis

Recently, emerging role of exosome (EXOSC) has been studied in WNT secretion and transportation by [34]. It has been found that exosomes play a critical role in morphogen signaling during embryonic development and cancer progression. In injured CNS, exosomes mediate mobilization of WNT10B to promote axonal regeneration as shown by [35]. [36] show the importance of exsosomes in WNT transportation. Emerging on these lines, we conducted a small two-way analysis of EXOSC components and WNT10B which were found to be down regulated in CRC cells after administration of ETC-1922159. Note that here, the interpretation of the rankings changes as the low numerical valued ranks (nearing to 1) are considered of high importance as they point to the synergistic down regulation after the drug administration. In line with the experiments, as ETC-1922159 a PORCN-WNT inhibitor block the transportation of WNTs, it might be that the affects of EXOSC components are also down regulated. These were rightly allocated with the low numerical valued in-silico ranks by the engine, thus pointing to the experimental down regulation in cells also. This confirmatory results also helps us in exploring the unknown combinations that might be prevailing synergistically when the WNT-EXOSC were up regulated before the administration of ETC-1922159 in CRC cells.

Table 8 shows the rankings of EXOSC family w.r.t WNT10B and vice versa. Followed by this is the unexplored combinatorial hypotheses in Table 9 generated from two-way analysis of the ranks in Table 8. On the left half of the Table 8, except for EXOSC7 - WNT10B, all other combinations of EXOSC family show high synergy with WNT10B. This is depicted by the low numerical valued ranks allocated by the search engine for EXOSC-2/3/5/6/8/9 with WNT10B, via majority voting across the ranking methods using laplace, linear and rbf kernels. This shows that EXOSC-2/3/5/6/8/9 had a critical role in the transport of WNT10B. On the right half of the same table, EXOSC-2/5/6/7/9 show synergistic affiliation with respect to WNT10B, via low numerical valued ranks. These are translated to graphical influences in Table 9. One can also interpret the results of the Table 9 graphically, with the following influences - • EXOSC w.r.t WNT10B with EXOSC-2/5/6/7/9 $<-$ WNT10B and • WNT10B w.r.t EXOSC with EXOSC-2/3/5/6/8/9 $->$ WNT10B. Further analyses of these combinations in wet lab might help biologists explore the deeper mechanism of exosome components and WNT10B in CRC cells.

3.1.6. CASP - WNT Cross Family Analysis

[37] show that a caspase-dependent pathway is involved in Wnt/$\beta$-catenin signaling promoted apoptosis in Bacillus Calmette-Guerin infected RAW264.7 macrophages. [38] have shown that WNT11 promotes cardiomyocyte development by caspase-mediated suppression of canonical WNT signals. Additionally, [39] show that Wnt5a and Wnt11 inhibit the canonical Wnt pathway and promote cardiac progenitor development via the Caspase-dependent degradation of AKT. These findings indicate probable interplay of Caspase and WNTs in various pathological cases. In mice, caspase-1 activation and IL-1$\mathsf{\beta }$ secretion together have shown to contribute to inflammatory condition of acute arthritis (see [40]). Recently, Caspase-3 inhibition has been found to be a therapeutic approach in colorectal cancer as shown by [41]. [42] also show synergistic role of Caspase-8 and Caspase-3 expressions as biomarkers in colorectal cancer. Family members belonging to each of the factors like CASP, WNT etc, might be involved synergistically in pathological case or otherwise. CASP and WNT members were found to be up regulated after the treatment of colorectal cancer cells with ETC-1922159.

Table 10 shows the rankings of CASP family w.r.t WNTs and vice versa. Followed by this is the unexplored combinatorial hypotheses in Table 11 generated from two-way analysis of the ranks in Table 10. On the first three tabular rows of the Table 10 show rankings of CASP family w.r.t WNT family. Here we present the possible interdependent WNT-CASP combinations that might be working synergistically in CRC cells. Considering CASP5 w.r.t WNTs, CASP5 - WNT2B show up regulated synergy through rankings of 2171 (laplace) and 2366 (linear). Considering CASP9 w.r.t WNTs, CASP9 - WNT-4/7B/9A show up regulated synergy through rankings of 2472 (laplace) and 2200 (linear) for CASP9 - WNT4; 2196 (laplace) and 1935 (linear) for CASP9 - WNT7B; and 1863 (laplace) and 2002 (linear) for CASP9 - WNT9A, respectively. Finally, considering CASP16 w.r.t WNTs, CASP16 - WNT4 showed up regulated synergy with rankings of 2070 (laplace) and 1783 (linear).

The next three tabular rows show rankings of WNT family w.r.t CASP family. W.r.t CASP4, WNT-7B/9A show promise of up regulation. These are reflected with rankings of 2479 (linear) and 1739 (rbf) for WNT7B - CASP4 and rankings of 2278 (linear) and 1939 (rbf) for WNT9A - CASP4, respectively. W.r.t CASP5, WNT-7B shows promise of up regulation. This is reflected with rankings of 2112 (laplace), 1919 (linear) and 2440 (rbf) for WNT7B - CASP5. W.r.t CASP7, WNT-2B/4/9A show promise of up regulation. These are reflected with rankings of 2505 (laplace) and 1891 (linear) for WNT2B - CASP7; rankings of 2456 (linear) and 2455 (rbf) for WNT4 - CASP7; and rankings of 2183 (laplace) and 1941 (linear) for WNT9A - CASP7, respectively. W.r.t CASP9, WNT-9A shows promise of up regulation. This is reflected with rankings of 2378 (laplace), 2396 (linear) and 2058 (rbf) for WNT9A - CASP9. W.r.t CASP10, WNT-4/9A show promise of up regulation. These are reflected with rankings of 1830 (laplace), 2229 (linear) and 1847 (rbf) for WNT4 - CASP10; and rankings of 2185 (laplace) and 1977 (linear) for WNT9A - CASP10, respectively. Finally, w.r.t CASP16, WNT-2B/4/9A show promise of up regulation. These are reflected with rankings of 2197 (laplace), 2489 (linear) and 1775 (rbf) for WNT2B - CASP16; rankings of 2508 (laplace), 1820 (linear) and 1867 (rbf) for WNT7B - CASP16; and rankings of 1943 (laplace) and 1839 (linear) for WNT9A - CASP16, respectively.

One can also interpret the results of the Table 11 graphically, with the following influences - • CASP w.r.t WNT with CASP5 $<-$ WNT2B; CASP9 $<-$ WNT-4/7B/9A; CASP16 $<-$ WNT4 and • WNT w.r.t CASP with. WNT-7B/9A $<-$ CASP4; WNT7B $<-$ CASP5; WNT-2B/4/9A $<-$ CASP7; WNT9A $<-$ CASP9; WNT-4/9A $<-$ CASP10; WNT-2B/7B/9A $<-$ CASP16.

3.1.7. TP53 - WNT Cross Family Analysis

[43] have shown that down regulation of $\beta$-catenin is activated by TP53. Wnt/$\beta$-catenin signaling is known to regulate the proliferation and differentiation of mesenchymal progenitor cells through the TP53 Pathway, as shown by [44]. [45] show that WNT activation by lithium abrogates TP53 mutation associated radiation resistance in medulloblastoma. In mouse cochlea, [46] show that WNT signaling activates TP53-induced glycolysis and apoptosis regulator and protects against cisplatin-induced spiral ganglion neuron damage. These range of interactions of TP53 with WNT points towards definite synergy. [47] show that TP53 protein regulates Hsp90 ATPase activity and thereby Wnt signaling by modulating Aha1 expression. Family members belonging to each of the factors like TP53, WNT etc, might be involved synergistically in pathological case or otherwise. TP53 and WNT members were found to be up regulated after the treatment of ETC-159 in colorectal cancer cells.

Table 12 contains rankings of TP53 w.r.t WNTs and vice versa. Followed by this is the unexplored combinatorial hypotheses in Table 13 generated from two-way analysis of the ranks in Table 12. On the left half of Table 12 are rankings of TP53 w.r.t WNTs and on the right half are the rankings of WNTs w.r.t TP53 family. Beginning with the left half, TP53I3 - WNT2B shows synergistic up regulation with rankings of 2056 (laplace) and 1712 (linear); TP53INP1 - WNT2B shows synergistic up regulation with rankings of 1805 (linear) and 2056 (rbf) and TP53BP2 - WNT9A shows synergistic up regulation with rankings of 2232 (linear) and 2143 (rbf). On the right half the table, TP53INP1 - WNT2B shows synergistic up regulation with rankings of 1853 (laplace) and 2089 (linear); TP53INP2 - WNT2B shows synergistic up regulation with rankings of 1723 (linear) and 2335 (rbf); TP53INP1 - WNT4 shows synergistic up regulation with rankings of 2414 (linear) and 2493 (rbf); TP53I3 - WNT7B shows synergistic up regulation with rankings of 1988 (laplace) and 2393 (rbf) and finally, TP53INP1 - WNT9A shows synergistic up regulation with rankings of 2045 (linear) and 2437 (rbf).

One can also interpret the results of the Table 11 graphically, with the following influences - • TP53 family w.r.t WNTs with TP53I3 $<-$ WNT2B; TP53INP1 $<-$ WNT2B and TP53BP2 $<-$ WNT9A; and • WNT family VS TP53 with TP53INP1 $->$ WNT2B; TP53INP2 $->$ WNT2B; TP53INP1 $->$ WNT4; TP53I3 $->$ WNT7B and TP53INP1 $->$ WNT9A.

3.1.8. BCL - WNT Cross Family Analysis

[48] observed that silencing Wnt2B by siRNA interference inhibits metastasis and enhances chemotherapy sensitivity in ovarian cancer. More specifically, [48] show that in the presence of Wnt2B siRNA treatment, the caspase-9/B-cell lymphoma 2 (BCL2)/B-cell lymphoma-xL (BCL-xL) pathway and the epithelial-mesenchymal transition/phosphorylated protein kinase B pathway were inhibited. [49] show that targeted disruption of the BCL9/$\beta$-catenin complex inhibits oncogenic WNT signaling. CDK1-mediated BCL9 phosphorylation inhibits clathrin to promote mitotic Wnt signaling as shown by [50]. These findings point to the existing synergy of BCL family with WNTs. Family members belonging to each of the factors like BCL, WNT etc, might be involved synergistically in pathological case or otherwise. BCL and WNT members were found to be up regulated after the treatment of ETC-159 in colorectal cancer cells.

Table 14 contains rankings of BCL w.r.t WNTs and vice versa. Followed by this is the unexplored combinatorial hypotheses in Table 15 generated from two-way analysis of the ranks in Table 14. On the left half of Table 14 are rankings of BCL w.r.t WNTs. WNT4 - BCL2L2 shows high ranking with 2364 (laplace) and 2042 (linear); WNT7B - BCL2L2 shows high ranking with 1877 (laplace) and 2456 (linear); WNT9A - BCL2L2 shows high ranking with 1877 (laplace) and 2447 (linear); WNT4 - BCL2L13 shows high ranking with 1938 (laplace), 2425 (linear) and 1900 (rbf); WNT7B - BCL2L13 shows high ranking with 1993 (linear) and 2284 (rbf) and WNT2B - BCL10 shows high ranking with 2321 (laplace) and 2023 (linear).

On the right side are rankings of WNTs w.r.t BCL. WNT7B - BCL2L1 shows high ranking with 2213 (laplace) and 2266 (linear); WNT7B - BCL2L2 shows high ranking with 2456 (laplace), 2512 (linear) and 2286 (rbf); WNT9A - BCL2L2 shows high ranking with 1868 (laplace) and 2333 (rbf); WNT9A - BCL2L13 shows high ranking with 1858 (laplace), 2422 (linear) and 1934 (rbf); WNT2B - BCL3 shows high ranking with 1846 (laplace), 2056 (linear) and 1896 (rbf); WNT4 - BCL6 shows high ranking with 2483 (laplace) and 2488 (linear); WNT7B - BCL6 shows high ranking with 1893 (laplace) and 2284 (linear); WNT9A - BCL6 shows high ranking with 2098 (linear) and 1905 (rbf); WNT2B - BCL9L shows high ranking with 1918 (laplace) and 1882 (rbf) and WNT4 - BCL9L shows high ranking with 2498 (linear) and 2509 (rbf);

One can also interpret the results of the Table 15 graphically, with the following influences - • BCL family w.r.t WNTs with WNT4 $->$ BCL2L2; WNT7B $->$ BCL2L2; WNT9A $->$ BCL2L2; WNT4 $->$ BCL2L13; WNT7B $->$ BCL2L13; WNT2B $->$ BCL10 and • WNT family w.r.t BCL with WNT7B $<-$ BCL2L1; WNT7B $<-$ BCL2L2; WNT9A $<-$ BCL2L2; WNT9A $<-$ BCL2L13; WNT2B $<-$ BCL3; WNT4 $<-$ BCL6; WNT7B $<-$ BCL6; WNT9A $<-$ BCL6; WNT2B $<-$ BCL9L; WNT4 $<-$ BCL9L.