Wikipedia network analysis of cancer interactions and world influence

We apply the Google matrix algorithms for analysis of interactions and influence of 37 cancer types, 203 cancer drugs and 195 world countries using the network of 5 416 537 English Wikipedia articles with all their directed hyperlinks. The PageRank algorithm provides a ranking of cancers which has 60% and 70% overlaps with the top 10 deadliest cancers extracted from World Health Organization GLOBOCAN 2018 and Global Burden of Diseases Study 2017, respectively. The recently developed reduced Google matrix algorithm gives networks of interactions between cancers, drugs and countries taking into account all direct and indirect links between these selected 435 entities. These reduced networks allow to obtain sensitivity of countries to specific cancers and drugs. The strongest links between cancers and drugs are in good agreement with the approved medical prescriptions of specific drugs to specific cancers. We argue that this analysis of knowledge accumulated in Wikipedia provides useful complementary global information about interdependencies between cancers, drugs and world countries.


Introduction
"Nearly every family in the world is touched by cancer, which is now responsible for almost one in six deaths globally" [1]. The number of new cancer cases in the world is steadily growing reaching 18.1 million projected for 2018 [2] with predicted new cases of 29.4 million for 2035 [3]. The detailed statistical analysis of new cases and mortality projected for 2018 is reported in [4]. Such statistical analysis is of primary importance for estimating the influence of cancer diseases on the world population. However, it requires significant efforts of research groups and medical teams all over the world such as consortia involved in the Global Burden of Diseases Study (GBD) [5] and the WHO GLOBOCAN reports [2].
Here we propose to probe the network of Wikipedia articles in order to infer specific interactions between cancer types and to measure world influence of cancers. Wikipedia can be seen as a global database of accumulated human knowledge with an immense variety of topics. Moreover the way Wikipedia articles are citing each other encodes scientific, social, historical, and many other aspects. In principle one should be able to extract from Wikipedia  REGOMAX analysis of countries are reported in [26,27]). The PageRank list of the 195 selected countries is available at [28]. Thus in total the reduced Google matrix selected number of nodes is N r = N cr + N d + N cn = 435. The inclusion of these three groups (cancer types, cancer drugs, and countries) in the reduced set of N r articles allows to investigate the interactions and influence of nodes inside group and between groups. The paper is composed as follows: the section "Description of data sets and methods" will present the May 2017 English Wikipedia network, introduce the Google matrix, the PageRank and CheiRank algorithms, and explain the construction of reduced Google matrices. In this section the node influence is defined through the PageRank ranking and the PageRank sensitivity. In the section "Results" we present the influence of cancer devoted pages in Wikipedia and extract a cancer ranking which is compared to cancer rankings extracted from GBD study [5] and GLOBOCAN [2] databases. We also use the reduced Google matrix to construct a reduced network of cancers and we determine the interaction of cancers with countries and cancer drugs. We corroborate the results obtained from the network structure of Wikipedia articles with various disease burden and/or epidemiological studies. To our knowledge, it is the first time that such correspondence is established. Finally we compare cancer prescriptions obtained from May 2017 English Wikipedia network analysis with approved medications reported in National Cancer Institute [25] and DrugBank [29]. The last section presents the conclusion of this research.

Network of English Wikipedia articles of 2017
We analyze the English language edition of Wikipedia collected in May 2017 (ENWIKI2017) [30] containing N = 5 416 537 articles (nodes) connected by N l = 122 232 932 directed hyperlinks  between articles (without self-citations). From this data set we extract the N cr = 37 types of cancers listed at [24]. From [25] we also collect names of drugs related to cancer diseases obtaining the list of N d = 203 drugs present at Wikipedia. The lists of 37 cancer types and 203 drugs are given in Table 1 and 2. This reduced set of N r = 240 nodes is illustrated in the inset of Fig 1. For global influence investigations, it is complemented by N cn = 195 world countries listed in [28]. Thus in total we have the reduced network of N r = N cr + N d + N cn = 435 � N nodes embedded in the global network with more than 5 millions nodes. All data sets are available at [28]. The present study complies with Wikimedia terms of use.

Google matrix construction rules
The construction rules of Google matrix G are described in detail in [6][7][8]. Thus the Google matrix G is built from the adjacency matrix A ij with elements 1 if article (node) j points to article (node) i and zero otherwise. The Google matrix elements have the standard form G ij = αS ij + (1 − α)/N [6][7][8], where S is the matrix of Markov transitions with elements S ij = A ij / k out (j). Here k out ðjÞ ¼ P N i¼1 A ij 6 ¼ 0 is the out-degree of node j (number of outgoing links) and S ij = 1/N if j has no outgoing links (dangling node). The parameter 0 < α < 1 is the damping factor. For a random surfer, jumping from one node to another, it gives the probability (1 − α) to jump to any node. Below we use the standard value α = 0.85 [7] noting that for the range 0.5 � α � 0.95 the results are not sensitive to α [7,8].
The right PageRank eigenvector of G is the solution of the equation GP = λP with the unit eigenvalue λ = 1. The PageRank components P(j) give positive probabilities to find a random surfer on a node j after an infinite journey (∑ j P(j) = 1). The numerical computation of P(j) is done efficiently with the PageRank algorithm described in [6,7].
The node influence is measured from the PageRank algorithm. We sort network nodes by decreasing PageRank probabilities. We assign K = 1 index to the node with maximal probability, i.e., the most central node according to PageRank algorithm, K = 2 index to the node with the second biggest probability, . . .A recursive definition of the PageRank algorithm can be given: a node is all the more influential as it is pointed by influential nodes.
It is also useful to consider the network with inverted direction of links. After links inversion A � ij ¼ A ji , the Google matrix G � is constructed within the same procedure with G � P � = P � . The matrix G � has its own PageRank vector P � called CheiRank [31] (see also [8,32]). Its probability values can be again ordered in a decreasing order with CheiRank index K � with highest P � (j) at K � = 1 and smallest at K � = N. The CheiRank algorithm measures the node diffusivity. A recursive definition of the CheiRank algorithm can also be given: a node is all the more diffusive as it is pointed by diffusive nodes.
On average, the high values of P(j) (P � (j)) correspond to nodes j with many ingoing (outgoing) links [8].
The PageRank order list of 37 cancers and 203 drugs is given in Table 3. In the global ENWIKI2017 network, countries are located on top PageRank positions (1. USA, 4. France, 5. Germany) so that cancers and drugs are located well below them since the first cancer type, i.e. Lung cancer, appears at 3 478th position, and the first cancer drug, i.e. Talc, appears at 22 177th position (see Fig 2). As expected cancer types have a more central position than cancer drugs. The network of 40 nodes and their direct links is shown in Fig 1 for the top 20 PageRank nodes of cancers and drugs (ordered separately for cancers and drugs). We see that already only for 40 nodes the network structure is rather complex. Here and below the networks are drawn with Cytoscape [33].

Reduced Google matrix algorithm
The details of REGOMAX method are described in [16,17,20]. It captures in the reduced Google matrix of size N r × N r the full contribution of direct and indirect pathways existing in  Table 3). The bulk of the other Wikipedia articles is not shown. Arrows symbolize hyperlinks between cancer and cancer drug articles in the global Wikipedia.
https://doi.org/10.1371/journal.pone.0222508.g001    Table 1 (Table 2) the full Google matrix between N r nodes of interest. The reduced Google matrix G R is such as G R P r = P r where P r is its associated PageRank probability vector. The PageRank probabilities P r (j) of the selected N r nodes are the same as for the global network with N nodes, up to a constant multiplicative factor taking into account that the sum of PageRank probabilities over N r nodes is unity. The computation of G R provides a decomposition into matrices that clearly distinguish direct from indirect interactions: G R = G rr + G pr + G qr [17]. Here G rr is the N r × N r submatrix of the N × N global Google matrix G encoding the direct links between the selected N r nodes. The G pr matrix is rather close to the matrix in which each column is given by the PageRank vector P r , ensuring that PageRank probabilities of G R are the same as for G (up to a constant multiplier). Thus G pr does not provide much more information about direct and indirect links between selected nodes than the usual Google matrix analysis described in the previous section. The component playing an interesting role is G qr , which takes into account all indirect links between selected nodes appearing due to multiple paths via the global network of N nodes (see [16,17]). The matrix G qr = G qrd + G qrnd has diagonal (G qrnd ) and nondiagonal (G qrnd ) parts. Thus G qrnd describes indirect interactions between nodes. The explicit formulas as well as the mathematical and numerical computation methods of all three components of G R are given in [16,17,20].
With the reduced Google matrix G R and its components we can analyze the PageRank sensitivity in respect to specific links between N r nodes. To measure the sensitivity of a country cn to a cancer cr we change the matrix element (G R ) cn,cr by a factor (1+δ) with δ � 1 and renormalize to unity the sum of the column elements associated with cancer cr, and we compute the logarithmic derivative of PageRank probability P(cn) associated to country cn: D(cr ! cn, cn) = d ln P(cn)/dδ (diagonal sensitivity). It is also possible to consider the nondiagonal (or indirect) sensitivity D(cr ! cn, cn 0 ) = d ln P(cn 0 )/dδ when the variation is done for the link from cr to cn and the derivative of PageRank probability is computed for another country cn 0 . Also instead of the link cr ! cn we can consider the link from a cancer cr to a drug d computing then the nondiagonal sensitivity of country cn 0 . The PageRank sensitivity approach, already used in [26,27], allows to measure the sensitivity of a node influence with respect to the change of a link intensity. Pragmatically it measures the increase or decrease of influence of a node A caused by an increase of the intensity of the link from a node B to a node C.

Cancer distribution on PageRank-CheiRank plane
The PageRank order of 37 cancers and 203 cancer drugs is given in Table 3. In the top 3 positions we find Lung, Breast, Leukemia cancers. Lung and Breast cancers have indeed the two highest incidences [2] and Leukemia is the most frequent type of cancer in children and young adults [34]. In general in the PageRank order of 240 cancers and drugs, cancers occupy predominantly the top positions. The first three drugs are Talc, Methotrexate, Thalidomide, taking positions 14, 20, 22. The top position of Talc among cancer drugs may be explained by its industrial use and also by both potential carcinogenic and anticancer effects [35]. Methotrexate can be used in the most frequent types of cancer but also in autoimmune diseases and for medical abortions [36]. The third position of Thalidomide among cancer drugs may be explained by its high potential for the treatment of cancers but also for its well-known teratogenic effect; this teratogenic effect may by itself contribute to its prominence in Wikipedia. It is also used for treatment of other diseases than cancers (tuberculosis, graft-versus-host disease,. . .) [37]. The list of these 240 articles in CheiRank order is also given in [28].
The distribution of selected articles on the global PageRank-CheiRank plane of the whole Wikipedia network with N = 5 416 537 nodes are shown in Fig 2. The top PageRank positions are taking by the world countries as discussed in [8,26] marked by gray open circles. Then there is a group of cancers (above K * 3 × 10 3 and K � * 10 4 ), marked by green points, followed by drugs (mostly above K * 10 4 and K � * 10 5 ), marked by gold points. There is a certain overlap between cancers and drugs on this plane but in global there is a clear separation between these two groups. As a comparison we also mark the positions of 230 infectious diseases by open blue circles. These 230 articles are studied in [27] in the frame of Wikipedia network analysis. The global PageRank list of 230 infectious diseases and 37 cancers is given in [28]. In this list Lung cancer is located at the 7th position. From Fig 2 we observe these two types of diseases occupy somewhat the same (K, K � ) region (mostly above K � * 10 5  The 240 cancer types and drugs placed on the plane of local PageRank indices K r 2 {1, . . ., 240} and CheiRank indices K � r 2 f1; . . . ; 240g is shown in Fig 3. We retrieve the fact that cancer types occupy the top positions in K r and in K � r . Indeed the first 14 most influent articles of this subset (K � 14), which appear to be devoted to cancer types, are also the most communicative with the exception of articles devoted to drugs Paclitaxel ðK r ¼ 24; [38] is a chemotherapy medication used to treat a wide range of cancer types e.g. Ovarian cancer, Breast cancer, Lung cancer, Pancreatic cancer, etc. Moreover Paclitaxel article cites Ovarian cancer article ðK r ¼ 10; K � r ¼ 1Þ which is a very communicative article since the Ovarian cancer article CheiRank index, K � = 29 317, is about one order magnitude lower than the CheiRank indexes, K � ≳ 10 5 , of the other 239 considered articles (see Fig 2). The wide applications of Paclitaxel and the citation of Ovarian cancer article explain the very good ranking of this cancer drug in the CheiRank scale. On the other hand, the K � r ¼ 2 rank of the Bicalutamide article (see Fig 3), devoted to an antiandrogen medication mainly used to treat Prostate cancer, is due to a very long article with a high density of intra-wiki citations [39]. Like the Paclitaxel article, the Bicalutamide article cites also the Ovarian cancer since this medication has already been tried for this cancer type [39].
The three most influent cancer drugs in ENWIKI2017 are Talc, K r = 14, which is used to prevent blood effusions, e.g., in Lung cancer or Ovarian cancer [35], Methotrexate, K r = 20, which is a chemotherapy agent used for the treatment Breast cancer, Leukemia, Lung cancer, Lymphoma, etc [36], and Thalidomide, K r = 22, which is a drug modulating the immune system used, e.g., for Multiple myeloma treatment [37]. Although Talc is widely used in chemical, pharmaceutical and food industries [35], its global PageRank position is nevertheless of the same order than the PageRank position of the second most influent cancer drug in Wikipedia, i.e., Methotrexate, which is a drug more specific to cancers [36].

Comparison of Wikipedia network analysis with GBD study 2017 and GLOBOCAN 2018 for cancer significance
We perform the comparison of cancer significance given by the GBD study 2017 [5], the GLO-BOCAN 2018 [2], and the Wikipedia network analysis. We extract the rankings of cancer types by the number of deaths in 2017 estimated by the 2017 GBD study [40] (see Table 4) and by the number of disability-adjusted life years (DALYs) estimated by the 2017 GBD study [41] (see Table 4). Also, we extract the rankings of cancer types by the number of deaths and by the number of new cases in 2018 estimated by the GLOBOCAN 2018 [4] (see Table 5). In Fig 4, we show the overlap of these 4 rankings with the extracted ranking of cancer types obtained from the ENWIKI2017 PageRanking (see bold items in Table 3). We observe that the ranking obtained from the Wikipedia network analysis provides a reliable cancer types ranking since its top 10 (top 20) shares about 70% (80%) similarity with GBD study data and GLOBOCAN data. The Wikipedia top 5 reaches even 80% similarity with top 5 cancer types extracted from the estimated number of new cases in 2018.

Reduced Google matrix of cancers and drugs
Let us consider now the subset of N r = 40 nodes composed of the first 20 cancers and the first 20 cancer drugs of the ENWIKI2017 PageRanking (Table 3). For this sub-network of interest illustrated in Fig 1, we perform the calculation of the reduced Google matrix G R and its components G rr , G pr and, G qr . From Fig 5, as expected, we observe that the G R matrix (top left panel) is dominated by the G pr component (bottom left panel) since W pr = 0.872 W R . The G pr component is of minor interest as it expresses again the relative PageRanking between the N r = 40 cancers and drugs already obtained and discussed in previous sections. The G rr (top right panel) gives the direct links between the considered cancers and drugs. Indeed, the G rr matrix is similar to the adjacency matrix A since there is a one-to-one correspondence between non zero entries of G rr and of A (for G rr by non zero entry we mean an entry greater than (1 − α)/N ' 2.8 × 10 −8 ). G qrnd (bottom right panel), the non diagonal part of G qr . We can consider that a link with a non zero entry in G qrnd and a zero entry in G rr (consequently also in A) is a hidden link. Below we use the non obvious components of G rr + G qrnd to draw the structure of reduced network.

Reduced network of cancers
We construct the reduced Google matrix associated to the set of N r = N cr + N cn = 232 Wikipedia articles constituted of N cr = 37 articles devoted to cancer types and of N cn = 195 articles devoted to countries. We consider the top 5 cancer types appearing in the ranking of May 2017 English Wikipedia using the PageRank algorithm which, according to Table 3, are 1 Lung cancer, 2 Breast cancer, 3 Leukemia, 4 Prostate cancer, 5 Colorectal cancer. Let us ordinate cancer types by their relative ranking in Table 3, cancer type cr i is consequently the ith most influent cancer type in May 2017 English Wikipedia. Using the reduced Google matrix, the component ðG rr þ G qrnd Þ cr i ;cr j , where i,j 2 {1,. . .,N cr }, gives the non obvious strength of the link pointing from the jth to the ith most influent cancer types. From each one the top 5 cancer types, {cr j } j2{1,. . .,5} , we select the two cancer types cr i 1 and cr i 2 , with i 1 ,i 2 2 {1,. . .,j − 1,j + 1,. . ., N cr }, to which cancer type cr j is preferentially linked ("friends"), i.e. those giving the two strongest ðG rr þ G qrnd Þ cr i ;cr j components. Around the main circle in Fig 6 (top panel) we first place the top 5 most influent cancer types in May 2017 English Wikipedia. Then we connect each Table 4 [40,41]. one of these cancer types to their two above defined cancer type friends. If these cancer types are not yet present in the network we add them in the vicinity of the cancer type pointing them. For each newly added cancer type we reiterate the same process until no new cancer type is added to the reduced network. The construction process of the reduced network of cancer ends at the 3rd iteration (see Fig 6, Table 4). In the reduced network of cancers showed in Fig 6 (top panel) we observe that the most influent cancer, i.e., Lung cancer is pointed from all the other cancer types with the exception of Hodgkin and Non-Hodgkin lymphomas. Also, Fig 6 (top panel) exhibits clearly a cluster of cancers (Colorectal, Stomach, and Liver cancers) affecting the digestive system, a cluster of cancers (Hodgkin and Non-Hodgkin lymphomas, and Leukemia) affecting blood, a loop interaction between Prostate and Breast cancers which are both linked to steroid hormone pathways and may be both treated with hormone therapy [42,43], loop interactions between Lung and Breast cancers and between Lung cancer and Melanoma affecting mainly the thoracic region. It is worth to note that although Leukemia article in May 2017 English Wikipedia does not cite any of the other articles devoted to cancer types (as an illustration the first half of the Leukemia column in G rr is filled with zero entries, see Fig 5 top right panel), we are able to infer In the reduced network of cancer, Fig 6 (top panel), we connect to each cancer types the two preferentially linked countries, i.e., for each cancer type cr, the two countries cn 1 and cn 2 giving the two highest value ðG rr þ G qrnd Þ cn;cr . We observe that cancers affecting digestive system point preferentially to Asian countries with the exception of Great Britain and Chile (Liver cancer points to Thailand and Saudi Arabia, Stomach cancer to Mongolia and Chile, Colorectal cancer to Philippines and Great Britain). This results are correlated to the fact that high mortality rates for Liver cancer are found in Asia (with the highest death rates for Eastern Asia [44]), and for Stomach cancer in Eastern Asia and South America [45,46]. In the other hand Colorectal cancer epidemiology clearly states [47] that the highest incidence rates are found for Western countries such as Great Britain. The appearance of Philippines pointed by Colorectal cancer is an artifact due to the mention in the corresponding 2017 Wikipedia article of Corazon Aquino, former president of the Philippines who was diagnosed with this cancer type. Blood cancer types points preferentially to African countries with the exception of Cambodia pointed by Hodgkin lymphoma. At first sight this results can appear surprising since these blood cancers are found worldwide with incidence rates highest for Western countries and  Table 3) and in the top j of the ranking of cancers by estimated number of worldwide deaths from GBD 2017 data [40] (black line, see Table 4), by estimation of disability-adjusted life years from GBD 2017 data [41] (black dashed line, Table 4), by estimated number of worldwide deaths from GLOBOCAN 2018 data [4] (red line, Table 5), and by estimated number of new cases from GLOBOCAN 2018 data [4] (red dashed line, Table 5). Only the black plain line is visible, where black plain line, red plain line and black dashed line overlap, e.g., from j = 1 to j = 5.

. List of cancer types ordered by the estimated number of deaths during the year 2017 (A) and by the estimated disability-adjusted life years (DALYs) for 2017 (B). Data extracted from GBD Study
https://doi.org/10.1371/journal.pone.0222508.g004 lowest for African countries [48]. In fact there is a Non-Hodgkin lymphoma, the Burkitt's lymphoma [49], which mainly affects children in malaria endemic region, i.e., Equatorial and Sub-Equatorial Africa and Eastern Asia. Countries pointed by blood cancer types, i.e., Liberia, Zambia, Cameroon, Gabon and Cambodia, belong to these regions. Let us note that these cancers and countries are connected through hidden links. Melanoma points to Australia, which is, with New Zealand [50], the country having the highest rate of Melanoma, and points to Peru, where nine 2400 years old mummies have been found with apparent signs of Melanoma   Table 3). These 5 cancers are symbolized by plain green nodes distributed around the central gray circle. We determine the two cancers to which each of these 5 cancers are preferentially linked according to (G rr + G qrnd ). If not among the top 5 cancers, a newly determined cancer is placed on a gray circle centered on the cancer from which it is linked. Then for each one of the newly added cancers we determine the two best cancers to which they are each linked, and so on. This process is stopped once no new cancers can be added, i.e. at the 3rd iteration (top panel) and 4th iteration (bottom panel). Also, at each iteration the two countries (drugs) to which each cancer are preferentially linked are placed on the gray circle centered on the cancer; see top panel (bottom panel). No new links are determined from the newly added countries or drugs. On top panel, countries are represented by ring shaped nodes (red for American countries, yellow for African countries, cyan for Asian countries, blue for European countries, and orange for Oceanian countries). On bottom panel, drugs are represented by plain gold nodes. The arrows represent the directed links between cancers and from cancers to countries or drugs (1st iteration: plain line; 2nd iteration: dashed line; 3rd iteration: dotted line for top panel and dashed-dotted line for bottom panel; 4th [50]. Prostate cancer points preferentially to Japan, due to its exceptional low incidence on Japanese population in Japan and abroad [51,52], to Nigeria, since it is believe that black population is particularly at risk [53]. Lung cancer points to Germany, where in 1929 it was shown for the first time a correlation between smoking and Lung cancer [54,55], and to Bhutan which adopted a complete smoking ban since 2005 [54]. Hidden link from Breast cancer to Republic of San Marino should be related to the fact that inhabitants of San Marino commemorate Saint Agatha, patroness of the Republic and of breast cancer patients [56]. Hidden link from Breast cancer to Cambodia is more difficult to interpret. Let us now consider the reduced Google matrix associated to N r = N cr + N d = 240 May 2017 English Wikipedia articles devoted to N cr = 37 cancer types and to N d = 203 cancer drugs. As above the reduced network of cancer can be constructed (Fig 6, bottom panel). The construction process ends at the 4th iteration. The main structure of reduced network of cancers is the same as the previous with some exceptions. Pancreatic cancer is added to the digestive system cancers cluster and via hidden links, Melanoma points now to Skin cancer which points to Breast cancer. Consequently we observe a new cluster of thoracic region cancers comprising Skin, Breast, Lung cancers and Melanoma. Let us connect to each cancer types the two preferentially linked cancer drugs, i.e., for each cancer type cr, the two cancer drugs d 1 and d 2 giving the two highest value ðG rr þ G qrnd Þ d;cr . Using DrugBank database [29], we easily check that indeed each drug is currently used to treat the cancer type to which it is connected. Also, closely connected cancer types share the same medication, e.g., Skin cancer and Melanoma are treated by Vemurafenib and Dabrafenib which are enzyme inhibitor of BRAF gene [57], Leukemia and Non-Hodgkin lymphoma are treated by the antibody Rituximab targeting B-lymphocyte antigen CD20 [58]. On the other hand non connected cancer types can in some cases share the same medication, the monoclonal antibody Trastuzumab typically used for Breast cancer is now also considered as a drug for Stomach cancer since these two cancer types overexpress the HER2 gene [59]. Let us note that hidden links connecting Non-Hodgkin lymphoma to Cyclophosphamide and Rituximab capture also a current medication reported in DrugBank database [29].
The reduced network of cancers shown in Fig 6 depict in a relevant manner interactions between cancers, cancer-country and cancer-drug interactions through Wikipedia.

World countries sensitivity to cancers
We consider the reduced Google matrix associated to the set of N r = N cr + N cn = 232 Wikipedia articles constituted of N cr = 37 articles devoted to cancer types and of N cn = 195 articles devoted to countries. We compute the PageRank sensitivity D(cr ! cn,cn), i.e., the infinitesimal rate of variation of PageRank probability P(cn) when the directed link cr ! cn, ðG R Þ cn;cr , is increased by an amount dðG R Þ cn;cr , where δ is an infinitesimal. Fig 7 shows the world distribution of PageRank sensitivity D(cr ! cn,cn) to Lung cancer. In Fig 7, color categories are obtained using the Jenks natural breaks classification method [60]. The most sensitive countries are, as discussed in the previous section, Bhutan and Germany mainly because these countries are directly cited in Wikipedia's Lung cancer article. Besides articles devoted to these two countries the others are not directly linked from the Lung cancer article and the results obtained in Fig 7 (top panel) is consistent with GLOBOCAN 2018 data [4]: apart Micronesia/Polynesia, the most affected countries, in term of incidence rates, are iteration: dotted line for bottom panel). Black arrows correspond to links existing in the adjacency matrix, i.e., direct links, and red arrows are purely hidden links absent from the adjacency matrix but present in the G qr component of the reduced Google matrix G R . These networks have been drawn with Cytoscape [33]. https://doi.org/10.1371/journal.pone.0222508.g006 Eastern Europe, Eastern Asia, Western Europe, and, Southern Europe for males, and, Northern America, Northern Europe, Western Europe, and, Australia/New Zealand for females. The less affected are African countries for both sexes. Let us note that although incidence rates are very high for males in Micronesia/Polynesia according to [4], this fact is not captured by Wikipedia since Nauru, Kiribati, Tuvalu, Marshall Islands are the less PageRank sensitive countries. This is certainly due to the fact that articles devoted to these sovereign states are among the worst ranked articles devoted to countries in the May 2017 English Wikipedia ranking using PageRank algorithm. Their respective ranks are Nauru K = 7085, Kiribati K = 7659, Tuvalu K = 6201, Marshall Islands K = 4549 to compare e.g. with USA K = 1, France K = 4, Germany K = 5, etc (see PageRank indices of countries in [28]).
As complementary information, sensitivities of countries to Breast cancer and to Leukemia are given in [28].
In order to investigate cancer-drug interactions it is also possible to represent sensitivity of countries to the variation of links from a cancer to a drug. As an illustration, Fig 8 shows countries PageRank sensitivities to variation of Lung cancer ! Bevacizumab link. We see that in this case the sensitivity of countries is significantly reduced comparing to the direct sensitivity influence of lung cancer on world countries shown in Fig 7. Since the influence of this link variation is indirect for countries it is rather difficult to recover due to what indirect links the influence for specific countries is bigger or smaller. Among the most affected European countries we find Lichtenstein, Great Britain, Iceland, Portugal and Croatia while Germany and the Czech Republic are mostly unaffected. Another example of sensitivity of countries to cancerdrug link variation is given in [28].

Interactions between cancers and drugs
Let us investigate interactions between cancers and drugs considering the subnetwork of N cr = 37 cancers (see Table 1) and of the first 37 cancer drugs appearing in the PageRank ordered list Table 3. We do not consider Talc here since it is widely used in not only pharmaceutical industries.
We consider the sensitivity of cancer to drugs via the computation of D(cr ! d,cr) presented in Fig 9. Although the PageRank sensitivity is computed using the logarithmic derivative of the PageRank, globally the most sensitives cancers are the ones with the highest PageRank probability, i.e., the ones with lowest PageRank indices K (see Fig 2 and [25] and DrugBank [29] databases report these possible drug-cancer associations. Let us consider directly the reduced Google matrix associated to the top 20 cancer types and top 20 cancer drugs according to May 2017 English Wikipedia PageRank list (Table 3). This reduced Google matrix G R and its G rr , G pr and G qrnd components are shown in Fig 5. For each cancer cr of the 20 most influent cancer types in May 2017 English Wikipedia let us determine the three most connected drugs d, i.e., the three drugs with the highest value of ðG rr þ G qrnd Þ d;cr . In Table 6 we show the May 2017 English Wikipedia prescription for each one of the top 20 cancer types. Most of the prescribed drugs are approved drugs for the considered  Table 3 and the first 37 drugs (d) listed in Table 2 (Talc has been removed as its article is too general).
https://doi.org/10.1371/journal.pone.0222508.g009 cancer types according to National Cancer Institute [25] and DrugBank [29]. Some of the Wikipedia proposed drugs are in fact subject of passed, ongoing or planned clinical trials. Only Dexamethasone is in fact not specific to Brain tumor since it is a corticosteroid used to treat inflammation in many medical conditions [61]. We observe that hidden links gives also accurate medication, see drugs associated to Non-Hodgkin lymphoma and Bladder cancer in Table 6.
Conversely for each cancer drug d of the 20 most influent cancer drugs in 2007 English Wikipedia we determine the three most connected cancer types cr, i.e., the three cancer types with the highest value of ðG rr þ G qrnd Þ cr;d . In Table 7 we show for which cancers a drug is prescribed according to May 2017 English Wikipedia. Again the results are globally in accordance  Table 3 and the first 37 drugs (d) listed in Table 2 (Talc has been removed as its article is too general). https://doi.org/10.1371/journal.pone.0222508.g010 with National Cancer Institute [25] and DrugBank [29] databases. We note that hidden links here correspond mainly to clinical trials, e.g., Imatinib is an approved drug for treatment of certain forms of Leukemia, but experiments were or will be done for Breast cancer and Prostate cancer.
It would be interesting to thoroughly study the most connected drugs and cancer types from the hidden contributions only, i.e., from ðG qrnd Þ cr;d or d;cr matrix elements only, in order to test the possible predictive power of the reduced Google matrix analysis of Wikipedia networks. This study is beyond the scope of the present work and will be considered in a subsequent one.

Conclusion
Using PageRank and CheiRank algorithms, we investigate global influences of 37 cancer types and 203 cancer drugs through the prism of Human knowledge encoded in the English edition of Wikipedia considered as a complex network. From the ranking of Wikipedia articles using PageRank algorithm we extract the ranking of the most influent cancers according to Wikipedia. This ranking is in good agreement with rankings, by either mortality rates or yearly new cases, extracted from WHO GLOBOCAN 2018 [2] and Global Burden of Diseases study 2017 [5] databases.  Table 3), we give the three strongest cancer ! drug links, i.e., for a given cancer type cr we select the three cancer drugs d with the highest values ðG rr þ G qr Þ d;cr . Drug with red background indicates a pure hidden cancer ! drug link, i.e., the cancer type article in Wikipedia does not refer directly to the drug. For each cancer ! drug link, the drug is followed by a ✔ mark if it is indeed prescribed for the cancer type according to National Cancer Institute [25] and/or DrugBank [29]; by a ▲ mark if the drug appears only as a subject of passed, ongoing or planned clinical trials reported for the cancer type in DrugBank; and by a ✘ mark otherwise.
The recently developed algorithm of the reduced Google matrix allows to construct a reduced network of cancers taking into account all the information aggregated in Wikipedia. This network exhibits direct and hidden links between the most influent cancers which form clusters of similar or related cancer types. The reduced Google matrix gives also countries or cancer drugs which are preferentially linked to the most influent cancers. Inferred relations between cancer types and countries obtained from Wikipedia network analysis are in accordance with global epidemiology literature. The PageRank sensitivity of countries to cancer types gives also a complementary tool corroborating epidemiological analysis. As far as we know, it is the first study highlighting correspondence between Wikipedia network analysis and disease burden or epidemiological studies. Inferred interactions between cancers and cancer drugs allows to determine drug prescriptions by Wikipedia for a specific cancer. These Wikipedia prescriptions appear to be compatible with approved medications reported in National Cancer Institute [25] and DrugBank [29] databases.
The reduced Google matrix algorithm allows to determine a clear and compact description of global influences and interactions of cancer types and cancer drugs integrating well documented medical aspects but also historical, and societal aspects, all encoded in the huge amount of knowledge aggregated in Wikipedia since 2001.  Table 3), we give the three strongest drug ! cancer links, i.e., for a given drug d we select the three cancer types cr with the highest values ðG rr þ G qr Þ cr;d . Cancer type with red background indicates a pure hidden drug ! cancer link, i.e., the drug article in Wikipedia does not refer directly to the cancer type. For each drug ! cancer link, the cancer type is followed by a ✔ mark if the drug is indeed prescribed for the cancer type according to National Cancer Institute [25] and/or DrugBank [29]; by a ▲ mark if the drug appears only as a subject of passed, ongoing or planned clinical trials reported for the cancer type in DrugBank; and by a ✘ mark otherwise.

Drug
1st cancer type 2nd cancer type 3rd cancer type