New Results

gpps: An ILP-based approach for inferring cancer progression with mutation losses from single cell data

Simone Ciccolella, Mauricio Soto Gomez, Murray Patterson, Gianluca Della Vedova, Iman Hajirasouliha, Paola Bonizzoni
doi: https://doi.org/10.1101/365635

Abstract

Motivation In recent years, the well-known Infinite Sites Assumption (ISA) has been a fundamental feature of computational methods devised for reconstructing tumor phylogenies and inferring cancer progression where mutations are accumulated through histories. However, some recent studies leveraging Single Cell Sequencing (SCS) techniques have shown evidence of mutation losses in several tumor samples [19], making the inference problem harder.

Results We present a new tool, gpps, that reconstructs a tumor phylogeny from single cell data, allowing each mutation to be lost at most a fixed number of times.

Availability The General Parsimony Phylogeny from Single cell (gpps) tool is open source and available at https://github.com/AlgoLab/gppf.

1 Introduction

Recent developments of targeted therapies for cancer patients rely on the accurate inference of the clonal evolution and progression of the particular cancer. As discussed in several recent studies, such as [24] and [32], understanding the order of accumulation and prevalence of somatic mutations during cancer progression can help better devise therapeutic strategies. Moreover, studying the evolutionary history of tumors can provide some insights on which mutations lead to drug resistance.

The most widely studied techniques for inferring cancer progression rely on data from next-generation bulk sequencing experiments. In these cases, we sample mixtures of cells that are not homogeneous from a mutational profile (i.e., which mutations appear in a cell) point of view. Moreover, we cannot easily distinguish between cells: the only information we can have is, for each mutation, the fraction of cells in a sample carrying such mutation. Recently, many computational approaches have been developed for the analysis of bulksequencing data with the purpose of inferring tumoral subclonal decomposition and reconstructing tumor phylogenies (trees) [3, 9, 15, 17, 22, 23, 25, 30, 31, 33], but almost all of them model a tumor progression as the accumulation of mutations under the Infinite Sites Assumption, that is recurrent mutations and mutation losses are not allowed.

Single Cell Sequencing (SCS) greatly improves the resolution of the data available, as it provides the set of mutations of each cell analyzed. However, this technique is currently expensive and not especially reliable, since it produces datasets with a high amount of noise that include allelic dropout (false negatives) and missing values, due to lack of read coverage, as well as false positive calls – although this event is much rarer. Another source of noise is due to doublets, that is signals originating from two separate cells which are erroneously inferred to originate from a single cell: we point out this latter problem is fading away and can be tackled computationally. Still, we need efficient methods that are able to cope with the kind of data that SCS techniques are currently producing, by taming the difficulties due to the noise in data.

Various methods have been developed for this purpose [16, 28, 34], some of them introducing a hybrid approach of combining both SCS and VAF data [20, 21, 26, 29]. As stated before, most of these methods rely on the Infinite Sites Assumption (ISA), which states that a mutation is acquired at most once in the phylogeny and is never lost. The ISA was introduced in [18]. This simplifying assumption also leads to a computationally tractable model of evolution [11] called the perfect phylogeny. Cancer progression, however, is a very fast and aggressive form of evolution with limited data supporting neutral evolution [6], with some studies showing rather the evidence of selection [2, 6] – something that is particularly true in tumor samples after a relapse [6, 8, 10], where the tumor has already been highly selected by the therapy targeted to destroy it. Thus, one would be expect that we must abandon the strict Infinite Sites Assumption in this setting, and indeed this is the case, as some recent studies are finding hints suggesting that the ISA does not always hold [2, 4, 19]. In [4], the authors find that large deletions on several branches of a tree can span a shared locus, thus a given mutation may be deleted independently multiple times. In [2], the authors show that in certain cases, homozygous deletions in cancer genomes can even provide a selective growth advantage. Each (independent) deletion of an acquired mutation takes us further away from the ISA. Some recent methods such as TRaIT [26] and SiFit [34] allow deletions of mutations.

The Dollo model [27] of evolution is designed exactly for some of the cases where a perfect phylogeny does not represent the actual data. More precisely, the Dollo model requires each mutation to be acquired exactly once in the entire history analyzed, while removing all restrictions on the number of times that a mutation can be lost. The Dollo model as well as the Dollo(k) variants, where each mutation can be lost at most k times, has been introduced recently in the literature on algorithmic approaches for tumor progression inference [3, 5]. Unfortunately, the Dollo model does not have the convenient computational tractability of the perfect phylogeny model [11], hence requiring more sophisticated algorithms.

In this paper we propose (gpps), a tool of the gppf family, that exploits an Integer Linear Programming (ILP) approach to infer a tumor progression that can include mutation losses, from single cell sequencing data.

Figure 1:
Figure 1:

Example of a binary matrix that does not allow a perfect phylogeny, since columns a and b are in conflict, i.e., the four gametes rule [11] does not hold. The tree represents one of the possible Dollo phylogenies that explain the matrix.

2 Tumor phylogeny reconstruction from single cell data

In the most abstract formulation, we can see the cancer progression reconstruction problem as a character-based phylogeny reconstruction problem [12] where each character represents the presence/absence of a specific mutation in a cell.

The input of the problem is an incomplete binary matrix I, where the entry I[c, m] = 0 indicates that the cell c does not have the mutation m, while I[c, m] = 1 indicates that the cell c has the mutation m. Finally, we denote with I[c, m] = ? where there is not enough information on the presence/absence of mutation m in cell c. We recall that uncertainty about the presence of a mutation in a cell is a consequence of insufficient coverage in the sequencing, hence it is unavoidable.

However, uncertainty is not the only issue in the sequencing process: the input matrix I also contains false positives and false negatives. We assume that these errors occur independently and uniformly across all the (known) entries of I. Namely, P denotes the predicted matrix, i.e., the binary matrix without missing values computed by the algorithm. In this case, α denotes the false negative rate and β denotes the false positive rate. In other words, for each pair (c, m), Embedded Image

Figure 2:
Figure 2:

Example of two Dollo phylogenies that explain the same binary matrix. It is important to notice that the ancestral order of mutations c, a and b is inverted but the two different trees can equally explain the input binary matrix. In fact, in a Dollo phylogeny the order of two mutations can be inverted and, thank to the introduction of deletions, they could both be correct solutions for a given input.

Our goal is to find a matrix P that (1) corresponds to a phylogeny on the set of cells, and (2) maximizes the the likelihood Embedded Image of the observed matrix I [16]. In other words, we want to find the phylogeny, as expressed by the matrix P, that maximizes the likelihood of the observed matrix I [16]. We point out that the values of the unknown entries of the input matrix do not factor into the objective function.

A phylogeny is a rooted labeled tree T, where the label set corresponds to the set of mutation gains and losses. The state S(x) of a leaf x in T is defined as the set of mutations that are acquired and not lost in the path from the root of T to x. We say that the tree T encodes a matrix P if there exists a mapping σ of the rows of P to the leaves of T such that for each row r of P, it follows that C(r) = S(σ(r)) where C(r) is the set of columns which are 1 in r, and σ(r) denotes the leaf of T associated with r through σ. In other words, in the tree T we assume that the cell c has been extracted from the subpopulation σ(c). See Figure 1 for example of a phylogeny a matrix that it encodes.

We can express the likelihood of the matrix P as in Equation 1 — since the involved probabilities are in [0,1] it is convenient to move to a (linear) log-likelihood maximization objective function of the form: Embedded Image

2.1 The evolution model

The Dollo parsimony rule can be interpreted as the impossibility of having an identical mutation in the evolutionary trajectory. This rule can be translated in the phylogeny tree model as the unique introduction of any single mutation but any number of deletions of this mutation.

From an algorithmic point of view, phylogeny reconstruction with a Dollo evolutionary model is an NP-complete problem [1, 7]. A hierarchical chain of restricted versions of the model can be obtained by bounding the number of deletions for each character. We denote as Dollo(k) the evolutionary model in which each mutation can be acquired exactly once and can be lost at most k times. In this way Dollo(0) and Dollo(1) correspond to the perfect and persistent phylogeny models, respectively. In the tree generation process for the Dollo(k) model (k > 0) we are required to augment a perfect phylogeny representing the cancer progression by adding nodes which represent the loss of a mutation, i.e., a node labeled Embedded Image, representing the l-th loss of mutation m. The state of the leaf x is the set of mutations m that, in the path from the root to x, have been acquired — the path has a vertex labeled m+ — but never lost — the path has no vertex labeled Embedded Image. We stress that, when deletions are introduced, the set of feasible phylogenies which represent a given solution is no longer unique as in the case of perfect phylogeny — See Figure 2 for an example.

Testing if an incomplete matrix I has a phylogeny under the Dollo(k) model has been attacked via ILP for k = 0 [14], k = 1 [13], and for general k [3]. We will exploit the latter formulation to describe an ILP approach for tumor phylogeny reconstruction from single cell data.

First, we recall that a well known characterization of perfect phylogenies states that a complete binary matrix M has a directed perfect phylogeny if and only if it has no conflicting pair of columns — two columns are in conflict if they contain all three configurations (0, 1), (1, 0), (1, 1) — inducing the so-called forbidden matrix (c.f. Figure 1).

The ILP formulation on incomplete matrices [14] essentially consists of introducing a binary variable for each missing entry, and describing a set of constraints towards the goal of minimizing the conflicting pairs.

To adapt this approach to persistent phylogenies [13], we need a property:

Proposition 1. [3] Let M be an incomplete binary matrix. Let Me be the (incomplete) matrix obtained from M as follows: for each entry M [i, j] we have k + 1 entries Me[i, j+] and Embedded Image (for 1 ≤ l ≤ k) such that (1) if M [i, j] = 1 then Me[i, j+] = 1 and Embedded Image = 0 for 1 ≤ l ≤ k, (2) if M [i, j] = 0 then the entries Me[i, j+], Embedded Imageare all missing. Then M has a completion M* that has a Dollo(k) phylogeny if and only if Me has a completion Embedded Image such that if M (i, j) = 0 then Embedded ImageSee Figure 3 for an illustration.

Our main contribution is combining the ILP formulation of [3] with the definition of tumor perfect phylogeny reconstruction from single cell data, to obtain an ILP approach for tumor phylogeny reconstruction from single cell data that incorporates mutation losses in the model.

Figure 3:
Figure 3:

An input matrix M (top left), a Dollo(2) completion Mc (center left) and its corresponding phylogeny tree T (top right). The MD(2) extended matrix (bottom left) and a completion for the MIDPP(MD(2), RD(2)) according to Proposition 1. In the tree, boldfaced character corresponds to changes between each node and its parent.

3 The ILP formulation

In this section we present our ILP formulation for the tumor phylogeny reconstruction from single cell data.

We recall that the input of the problem is an incomplete matrix I represented as a set of binary variables I(c, m) such that I(c, m) = 1 if cell c has (according to the input data) the mutation m, while I(c, m) = 0 if cell c does not have (according to the input data) the mutation m. Notice that the input data is incomplete, hence we can have pairs (c, m) such that the variable I(c, m) does not exist.

The variables P (c, m+) and Embedded Image encode the extended matrix that we want to compute and that will satisfy Proposition 1. Differently from the variable I(.,.), for each pair (c, m), all variables P (c, m+) and Embedded Imageexist.

We introduce some auxiliary variables that help in making the ILP formulation easier to read. The binary variables F (c, m) indicates if, in the predicted matrix, the cell c has the mutation m. By Proposition 1, F (c, m) = 1 if and only if P (c, m+) = 1 and all Embedded Imageare equal to zero. Moreover, the real variables w(c, m) represent the probability of P (c, m) given I(c, m) — the formula of the actual values depends on the possible cases, that is if we have a true positive, a true negative, a false positive, and a false negative.

To establish if two columns are in conflict, we introduce the final binary variables B(p, q, a, b), which are defined for each pair of columns (p, q) and for each possible pair of values (a, b)∈{(0, 1), (1, 0), (1, 1)}. More precisely, B(p, q, a, b) indicates if for the pair (p, q) of columns there exists a cell c where P (c, p) = a and P (c, q) = b. Notice that two columns p and q are conflicting iff B(p, q, 0, 1) + B(p, q, 1, 0) + B(p, q, 1, 1) = 3. We are now ready to introduce our ILP formulation, where we use C to denote the set of cells (i.e., the rows of the input matrix I), M to denote the mutations (i.e., the columns of I), and M* to denote the set of possible mutation gains or losses.Embedded Image Embedded Image Embedded Image Embedded Image Embedded Image Embedded Image Embedded Image

The total number of variables and constraints in the formulation are O(nm + m2) and O(nm2) respectively.

3.1 ILP implementation: gpps

Our approach has been implemented with a Python program called gpps. The code, data and scripts used in our experimental analysis is available at https://github.com/AlgoLab/gppf. The algorithm receives as input a frequency matrix F, the evolution model (persistent, Dollo(k), Camin-Sokal(k)) to be considered and the maximum number of clones in the clonal matrix (expressed as the percentage of the total number of mutations). The program generates the ILP formulation which is fed to an ILP solver in order to get the optimal solution.

In our experiments we have used Gurobi 6.5.2 as the ILP solver. Moreover, from the computed solution the program can construct the corresponding solution tree, provided that feasible solution has been found. Additionally, we have introduced a timeout on the running time, since the generated ILP problem could be large and its resolution could require a considerable amount of time. We exploit the fact that Gurobi can be halted at any time and it returns the best feasible solution computed so far. Hence, imposing a timeout allows the ILP solver to compute a solution with a small total error.

Acknowledgement

SC acknowledges a Mobility Exchange Fellowship from the University of MilanoBicocca. Part of this work has been done during SC visit at Weill Cornell. This work was also supported by start up funds (Weill Cornell Medicine) to IH.

References

  1. [1].↵
    1. Zvi Galil and
    2. Esko Ukkonen
    Craig Benham, Sampath Kannan, and Tandy Warnow. Of chicken teeth and mouse eyes, or generalized character compatibility. In Zvi Galil and Esko Ukkonen, editors, Combinatorial Pattern Matching, pages 17–26, Berlin, Heidelberg, 1995. Springer Berlin Heidelberg.
  2. [2].↵
    Graham R. Bignell, Chris D. Greenman, Helen Davies, Adam P. Butler, Sarah Edkins, Jenny M. Andrews, Gemma Buck, Lina Chen, David Beare, Calli Latimer, Sara Widaa, Jonathon Hinton, Ciara Fahey, Beiyuan Fu, Sajani Swamy, Gillian L. Dalgliesh, Bin T. Teh, Panos Deloukas, Fengtang Yang, Peter J. Campbell, P. Andrew Futreal, and Michael R. Stratton. Signatures of mutation and selection in the cancer genome. Nature, 463:893–898, 2010.
    OpenUrlCrossRefPubMedWeb of Science
  3. [3].↵
    Paola Bonizzoni, Simone Ciccolella, Gianluca Della Vedova, and Mauricio Soto. Beyond perfect phylogeny: Multisample phylogeny reconstruction via ilp. In Proceedings of the 8th ACM International Conference on Bioinformatics, Computational Biology,and Health Informatics, ACM-BCB ′17, pages 1–10, New York, NY, USA, 2017. ACM.
  4. [4].↵
    David Brown, Dominiek Smeets, Borbála Székely, Denis Larsimont, A. Marcell Szász, Pierre-Yves Adnet, Françoise Rothé, Ghizlane Rouas, Zsófia I. Nagy, Zsófia Faragó, Anna-Mária Tokés, Magdolna Dank, Gyöngyvér Szentmártoni, Nóra Udvarhelyi, Gabriele Zoppoli, Lajos Pusztai, Martine Piccart, Janina Kulka, Diether Lambrechts, Christos Sotiriou, and Christine Desmedt. Phylogenetic analysis of metastatic progression in breast cancer using somatic mutations and copy number aberrations. Nature Communications, 8:14944 EP –, Apr 2017. Article.
    OpenUrl
  5. [5].↵
    Simone Ciccolella, Mauricio Soto Gomez, Murray Patterson, Gianluca Della Vedova, Iman Hajirasouliha, and Paola Bonizzoni. Inferring cancer progression from single cell sequencing while allowing loss of mutations. bioRxiv, 2018.
  6. [6].↵
    Alexander Davis, Ruli Gao, and Nicholas Navin. Tumor evolution: Linear, branching, neutral or punctuated? Biochimica et Biophysica Acta (BBA) - Reviews on Cancer, 1867(2):151–161, 2017. Evolutionary principles - heterogeneity in cancer?
    OpenUrl
  7. [7].↵
    William H.E. Day, David S. Johnson, and David Sankoff. The computational complexity of inferring rooted phylogenies by parsimony. Mathematical Biosciences, 81(1):33–42, 1986.
    OpenUrlCrossRefWeb of Science
  8. [8].↵
    L Ding, TJ Ley, DE Larson, CA Miller, DC Koboldt, JS Welch, JK Ritchey, MA Young, T Lamprecht, MD McLellan, et al. Clonal evolution in relapsed acute myeloid leukaemia revealed by whole-genome sequencing. Nature, 481:506–10, 2012.
    OpenUrlCrossRefPubMedWeb of Science
  9. [9].↵
    Mohammed El-Kebir, Gryte Satas, Layla Oesper, and Benjamin J. Raphael. Inferring the mutational history of a tumor using multi-state perfect phylogeny mixtures. Cell Systems, 3(1):43–53, 2018/01/25 XXXX.
    OpenUrl
  10. [10].↵
    RJ Gillies, D Verduzco, and Gatenby RA. Evolutionary dynamics of carcinogenesis and why targeted therapy does not work. Nat Rev Cancer, 12:487–93, 2012.
    OpenUrlCrossRefPubMedWeb of Science
  11. [11].↵
    Dan Gusfield. Efficient algorithms for inferring evolutionary trees. Networks, 21:19–28, 1991.
    OpenUrlCrossRef
  12. [12].↵
    Dan Gusfield. Algorithms on Strings, Trees and Sequences: Computer Science and Computational Biology. Cambridge University Press, Cambridge, 1997.
  13. [13].↵
    Dan Gusfield. Persistent phylogeny: a galled-tree and integer linear programming approach. In Proceedings of the 6th ACM Conference on Bioinformatics, Computational Biology and Health Informatics, pages 443–451. ACM, 2015.
  14. [14].↵
    1. Guohui Lin
    Dan Gusfield, Yelena Frid, and Dan Brown. Integer programming formulations and computations solving phylogenetic and population genetic problems with missing or genotypic data. In Guohui Lin, editor, Computing and Combinatorics: 13th Annual International Conference, COCOON 2007, Banff, Canada, July 16-19, 2007. Proceedings, pages 51–64. Springer Berlin Heidelberg, Berlin, Heidelberg, 2007.
  15. [15].↵
    Iman Hajirasouliha, Ahmad Mahmoody, and Benjamin J. Raphael. A combinatorial approach for analyzing intra-tumor heterogeneity from high-throughput sequencing data. Bioinformatics, 30(12):i78–i86, Jun 2014.
    OpenUrlCrossRefPubMed
  16. [16].↵
    Katharina Jahn, Jack Kuipers, and Niko Beerenwinkel. Tree inference for single-cell data. Genome Biology, 17(1):86, May 2016.
    OpenUrlCrossRef
  17. [17].↵
    Wei Jiao, Shankar Vembu, Amit G. Deshwar, Lincoln Stein, and Quaid Morris. Inferring clonal evolution of tumors from single nucleotide somatic mutations. BMC Bioinformatics, 15(1):35, 2014.
    OpenUrlCrossRefPubMed
  18. [18].↵
    Motoo Kimura. The number of heterozygous nucleotide sites maintained in a finite population due to steady flux of mutations. Genetics, 61(4):893–903, 1969.
    OpenUrlFREE Full Text
  19. [19].↵
    Jack Kuipers, Katharina Jahn, Benjamin J. Raphael, and Niko Beeren-winkel. Single-cell sequencing data reveal widespread recurrence and loss of mutational hits in the life histories of tumors. Genome Research, 2017.
  20. [20].↵
    Salem Malikic, Simone Ciccolella, Farid Rashidi Mehrabadi, Camir Ricketts, Md. Khaledur Rahman, Ehsan Haghshenas, Daniel Seidman, Faraz Hach, Iman Hajirasouliha, and S. Cenk Sahinalp. Phiscs - a combinatorial approach for sub-perfect tumor phylogeny reconstruction via integrative use of single cell and bulk sequencing data. bioRxiv, 2018.
  21. [21].↵
    Salem Malikic, Katharina Jahn, Jack Kuipers, Cenk Sahinalp, and Niko Beerenwinkel. Integrative inference of subclonal tumour evolution from single-cell and bulk sequencing data. bioRxiv, 2017.
  22. [22].↵
    Salem Malikic, Andrew W. McPherson, Nilgun Donmez, and Cenk S. Sahinalp. Clonality inference in multiple tumor samples using phylogeny. Bioinformatics, 31(9):1349–1356, 2015.
    OpenUrlCrossRefPubMed
  23. [23].↵
    Francesco Marass, Florent Mouliere, Ke Yuan, Nitzan Rosenfeld, and Florian Markowetz. A phylogenetic latent feature model for clonal deconvolution. Ann. Appl. Stat., 10(4):2377–2404, 12 2016.
    OpenUrl
  24. [24].↵
    A. Sorana Morrissy and Livia et al. Garzia. Divergent clonal selection dominates medulloblastoma at recurrence. Nature, 529:351 EP –, 01 2016.
    OpenUrl
  25. [25].↵
    Victoria Popic, Raheleh Salari, Iman Hajirasouliha, Dorna Kashef-Haghighi, Robert B. West, and Serafim Batzoglou. Fast and scalable inference of multi-sample cancer lineages. Genome Biol, 16(1):91, May 2015. 647[PII].
    OpenUrlCrossRefPubMed
  26. [26].↵
    Daniele Ramazzotti, Alex Graudenzi, Luca De Sano, Marco Antoniotti, and Giulio Caravagna. Learning mutational graphs of individual tumor evolution from multi-sample sequencing data. bioRxiv, 2017.
  27. [27].↵
    I. Rogozin, Y. Wolf, V. Babenko, and E Koonin. Dollo parsimony and the reconstruction of genome evolution. Oxford University Press.
  28. [28].↵
    Edith M. Ross and Florian Markowetz. Onconem: inferring tumor evolution from single-cell sequencing data. Genome Biology, 17(1):69, Apr 2016.
    OpenUrlCrossRef
  29. [29].↵
    Sohrab Salehi, Adi Steif, Andrew Roth, Samuel Aparicio, Alexandre Bouchard-Côté, and Sohrab P. Shah. ddclone: joint statistical inference of clonal populations from single cell and bulk tumour sequencing data. Genome Biology, 18(1):44, Mar 2017.
    OpenUrl
  30. [30].↵
    Gryte Satas and Benjamin J. Raphael. Tumor phylogeny inference using tree-constrained importance sampling. Bioinformatics, 33(14):i152–i160, 2017.
    OpenUrlCrossRef
  31. [31].↵
    Francesco Strino, Fabio Parisi, Mariann Micsinai, and Yuval Kluger. Trap: a tree approach for fingerprinting subclonal tumor composition. Nucleic Acids Res, 41(17):e165–e165, Sep 2013. gkt641[PII].
    OpenUrlCrossRefPubMed
  32. [32].↵
    Jiguang Wang, Emanuela Cazzato, Erik Ladewig, Veronique Frattini, Daniel I S Rosenbloom, Sakellarios Zairis, Francesco Abate, Zhaoqi Liu, Oliver Elliott, Yong-Jae Shin, Jin-Ku Lee, In-Hee Lee, Woong-Yang Park, Marica Eoli, Andrew J Blumberg, Anna Lasorella, Do-Hyun Nam, Gaetano Finocchiaro, Antonio Iavarone, and Raul Rabadan. Clonal evolution of glioblastoma under therapy. Nature Genetics, 48:768 EP –, 06 2016.
    OpenUrl
  33. [33].↵
    Ke Yuan, Thomas Sakoparnig, Florian Markowetz, and Niko Beerenwinkel. Bitphylogeny: a probabilistic framework for reconstructing intra-tumor phylogenies. Genome Biology, 16(1):36, Feb 2015.
    OpenUrlCrossRefPubMed
  34. [34].↵
    Hamim Zafar, Anthony Tzen, Nicholas Navin, Ken Chen, and Luay Nakhleh. Sifit: inferring tumor trees from single-cell sequencing data under finite-sites models. Genome Biology, 18(1):178, Sep 2017.
    OpenUrl
Back to top
Posted September 04, 2018.
Download PDF
Email
gpps: An ILP-based approach for inferring cancer progression with mutation losses from single cell data
Simone Ciccolella, Mauricio Soto Gomez, Murray Patterson, Gianluca Della Vedova, Iman Hajirasouliha, Paola Bonizzoni
bioRxiv 365635; doi: https://doi.org/10.1101/365635
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area