Overview of Overview of Complex Networks Outline Complex Networks Complex Networks Principles of Complex Systems Basic definitions Basic definitions Basic definitions Examples of Examples of Course CSYS/MATH 300, Fall, 2009 Complex Networks Complex Networks Properties of Examples of Complex Networks Properties of Complex Networks Complex Networks
Nutshell Nutshell Prof. Peter Dodds Properties of Complex Networks Basic models of Basic models of complex networks complex networks Dept. of Mathematics & Statistics Generalized random Generalized random networks Nutshell networks Center for Complex Systems :: Vermont Advanced Computing Center Scale-free networks Scale-free networks University of Vermont Small-world networks Small-world networks Generalized affiliation Generalized affiliation networks Basic models of complex networks networks References Generalized random networks References Scale-free networks Small-world networks Generalized affiliation networks
References Frame 1/122 Frame 2/122 Licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License.
Overview of Overview of net•work |ˈnetˌwərk| Complex Networks Thesaurus deliciousness: Complex Networks noun 1 an arrangement of intersecting horizontal and vertical lines. Basic definitions Basic definitions • a complex system of roads, railroads, or other transportation routes : a network of railroads. Examples of Examples of Complex Networks Complex Networks 2 a group or system of interconnected people or things : a trade network. • Properties of Properties of a group of people who exchange information, contacts, and Complex Networks Complex Networks experience for professional or social purposes : a support network. Nutshell network Nutshell • a group of broadcasting stations that connect for the simultaneous noun broadcast of a program : the introduction of a second TV network | [as adj. ] Basic models of Basic models of complex networks WEB complex networks network television. Generalized random 1 a network of arteries , lattice, net, matrix, mesh, Generalized random • a number of interconnected computers, machines, or operations : networks networks Scale-free networks crisscross, grid, reticulum, reticulation; Anatomy plexus. Scale-free networks specialized computers that manage multiple outside connections to a network | a Small-world networks Small-world networks Generalized affiliation 2 a network of lanes MAZE, labyrinth, warren, tangle. Generalized affiliation local cellular phone network. networks networks • a system of connected electrical conductors. 3 a network of friends SYSTEM, complex, nexus, web, References References webwork. verb [ trans. ] connect as or operate with a network : the stock exchanges have proven to be resourceful in networking these deals. • link (machines, esp. computers) to operate interactively : [as adj. ] ( networked) networked workstations. • [ intrans. ] [often as n. ] ( networking) interact with other people to exchange information and develop contacts, esp. to further one's career : the skills of networking, bargaining, and negotiation. Frame 3/122 Frame 4/122
DERIVATIVES net•work•a•ble adjective Overview of Overview of Ancestry: Complex Networks Ancestry: Complex Networks
Basic definitions Basic definitions
Examples of Examples of From Keith Briggs’s excellent Complex Networks First known use: Geneva Bible, 1560 Complex Networks Properties of Properties of etymological investigation: () Complex Networks ‘And thou shalt make unto it a grate like networke of Complex Networks Nutshell brass (Exodus xxvii 4).’ Nutshell Basic models of Basic models of complex networks complex networks Generalized random From the OED via Briggs: Generalized random networks networks Scale-free networks Scale-free networks Small-world networks I 1658–: reticulate structures in animals Small-world networks Generalized affiliation Generalized affiliation I Opus reticulatum: networks networks I 1839–: rivers and canals References References I A Latin origin? I 1869–: railways
I 1883–: distribution network of electrical cables
[http://serialconsign.com/2007/11/we-put-net-network] I 1914–: wireless broadcasting networks
Frame 5/122 Frame 6/122
Keith Briggs: : Etymology of `network' 6/6/09 9:45 PM gescylde hálgan nette (with a net-work of clouds), Cd. Th. 182, Ii; Exod. 74. [Goth, nati: O. Sax. netti, (fisk-)net: O. Frs. nette: Icel. net; gen. pl. netja : O. H. Ger. nezzi rete.] v. æl-, boge-, breóst-, deór-, drag-, feng-, fisc-, fleóh-, here-, bring-, inwit-, mycg-, searo-, wæl-nett, and next word. Overview of Overview of Ancestry: Complex Networks Key Observation: Complex Networks nette, an; f. The net-like caul :-- Nette (under the heading de membris hominum) disceptum i. reticulum (cf. hoc reticulum, pinguedo circa jecur [fat around liver], 704, 7), Wülck. Gl. 293,6. Nettae oligia, 35, 34. Nytte obligia [binding, bandage], Wrt. Voc. i. 45, 18. Nette, ii. 63, 39 : disceptum, 26,19. [Icel. netja the caul: cf. O. H. Ger. nezzi adeps [fat] intestini; pl. intestina.] v. neta. Basic definitions I Many complex systems Basic definitions Net has many cognatesNet in Germanicand languages and Work a probable one in are Latin. The venerableGermanic words are old words: of neuter gender (except for nót which is feminine); the Latin word is feminine. Note that the normal Latin word for a net is rete, although there may have also been a Vulgar Latin word *tragina for a Examples of can be viewed as complex networks Examples of dragnet. (King Ælfréd, 888) Complex Networks Complex Networks language word I ‘Net’ first used to mean spider web . of physical or abstract interactions. Gothic nati Properties of Properties of Old net(t), ‘Work’ appear to have long meant purposeful action. Complex Networks Complex Networks English netti I I Opens door to mathematical and numerical analysis. Middle net Nutshell Nutshell English Keith Briggs: : Etymology of `network' 6/6/09 9:45 PM Modern net Dominant approach of last decade of a English Basic models of I Basic models of Modern net complex networks complex networks Dutch theoretical-physics/stat-mechish flavor. Generalized random Generalized random Old High nezzi German networks networks Middle Scale-free networks I Mindboggling amount of work published on complex Scale-free networks High netze Small-world networks Small-world networks German Modern Generalized affiliation networks since 1998... Generalized affiliation High Netz networks networks German Old nette References References Frisian I ... largely due to your typical theoretical physicist: Old netti Saxon net Old (netja Norse v.), nót sources Modern I Piranha physicus net Icelandic Bosworth, Joseph: (Northcote Toller, T. ed.) An Anglo-Saxon Dictionary. Dictionary, Supplement and Modern I ‘Network’ = something builtAddenda based Oxford University on Press, 1st the ed. n.d., ideareprinted 1983, of 1966, 1972. [Abbreviations: Wrt. Voc. i: nät Swedish T. Wright, A volume of vocabularies, 1857. Quoted by page and number of gloss. Wrt. Voc. ii: T. Wright, I A second volume of vocabularies, 1873. Quoted by page and line. Ælfc. Gr: Ælfric's Grammar. Ps. Spl.: Hunt in packs. Latin nassa natural, flexible lattice or web.Psalterium Davidis Latino-Saxonicum vetus. Lk. Skt.: The gospel according to St. Luke. Mt. Kmbl.: The gospel according to St. Matthew in Anglo-Saxon and Northumbrian versions. ed. J. M. Kemble. Jn. Skt.: The Gothic word occurs exactly seven times in the bible. It takes the form natja in the dative. Because The gospel according to St. John. Som.: Dictionarium Saxonico-Latino-Anglicum, by E. Somner, Oxford Gothic is the earliest record of Germanic (from about 350), it is worth looking at all these occurrences: 1659. Coll. Monast. Th.: Colloquium ad pueros linguae Latinae locutione exercendos ad Ælfrico I Feast on new and interesting ideas I c.f., ironwork, stonework, fretwork.compilatum. Homl. Th.: The homilies of Ælfric. Wülck. Gl.: Anglo-Saxon and Old English vocabularies, by Thomas Wright, edited by R. P. Wülcker, London 1884. Cd. Th.: Cædmon's metrical paraphrase, by B. Thorpe, London 1832.] Frame 7/122 (see chaos, cellular automata, ...) Frame 8/122 http://keithbriggs.info/network.html Page 2 of 4 Stamm, Friedrich Ludwig: Friedrich Ludwig Stamm's Ulfilas oder die uns erhaltenen Denkmäler der gotischen Sprache: Text, Wörterbuch und Grammatik Ungekürzte Neuauflage, Nachdruck der Ausgabe aus dem Jahre 1872 [Essen]: Magnus-Verlag, [1984]. ISBN: 3884001655
This page was last modified 2005 Sep 26 (Tuesday) 09:52 by
http://keithbriggs.info/network.html Page 4 of 4 Overview of Overview of Popularity (according to ISI) Complex Networks Popularity according to books: Complex Networks
Basic definitions Basic definitions “Collective dynamics of ‘small-world’ networks” [28] Examples of Examples of Complex Networks Complex Networks Watts and Strogatz Properties of Properties of I Complex Networks The Tipping Point: How Little Things can Complex Networks Nature, 1998 Nutshell make a Big Difference—Malcolm Gladwell [12] Nutshell Basic models of Basic models of ≈ (as of June 5, 2009) I 3752 citations complex networks complex networks Generalized random Generalized random I Over 1100 citations in 2008 alone. networks networks Scale-free networks Scale-free networks Small-world networks Small-world networks Generalized affiliation Generalized affiliation “Emergence of scaling in random networks” [3] networks networks References References I Barabási and Albert Nexus: Small Worlds and the Groundbreaking Science, 1999 Science of Networks—Mark Buchanan
I ≈ 3860 citations (as of June 5, 2009)
I Over 1100 citations in 2008 alone.
Frame 9/122 Frame 10/122
Overview of Overview of Popularity according to books: Complex Networks Numerous others: Complex Networks
Complex Social Networks—F. Vega-Redondo [25] Basic definitions I Basic definitions Examples of I Fractal River Basins: Chance and Self-Organization—I. Examples of Complex Networks Complex Networks Rodríguez-Iturbe and A. Rinaldo [20] Properties of Properties of Linked: How Everything Is Connected to Complex Networks Complex Networks I Random Graph Dynamics—R. Durette Everything Else and What It Nutshell Nutshell Means—Albert-Laszlo Barabási Basic models of I Scale-Free Networks—Guido Caldarelli Basic models of complex networks complex networks Generalized random Generalized random networks I Evolution and Structure of the Internet: A Statistical networks Scale-free networks Physics Approach—Romu Pastor-Satorras and Scale-free networks Small-world networks Small-world networks Generalized affiliation Alessandro Vespignani Generalized affiliation networks networks References I Complex Graphs and Networks—Fan Chung References Six Degrees: The Science of a Connected Age—Duncan Watts [26] I Social Network Analysis—Stanley Wasserman and Kathleen Faust
I Handbook of Graphs and Networks—Eds: Stefan Bornholdt and H. G. Schuster [6] Evolution of Networks—S. N. Dorogovtsev and J. F. F. Frame 11/122 I Frame 12/122 Mendes [11] Overview of Overview of More observations Complex Networks More observations Complex Networks
Basic definitions Basic definitions
Examples of Examples of Complex Networks I Web-scale data sets can be overly exciting. Complex Networks I But surely networks aren’t new... Properties of Properties of Complex Networks Complex Networks I Graph theory is well established... Witness: Nutshell Nutshell I Study of social networks started in the 1930’s... Basic models of I The End of Theory: The Data Deluge Makes the Basic models of complex networks complex networks I So why all this ‘new’ research on networks? Generalized random Scientific Theory Obsolete (Anderson, Wired) ( ) Generalized random networks networks Scale-free networks Scale-free networks I Answer: Oodles of Easily Accessible Data. Small-world networks I “The Unreasonable Effectiveness of Data,” Small-world networks Generalized affiliation [13] Generalized affiliation networks Halevy et al. . networks I We can now inform (alas) our theories with a much more measurable reality.∗ References References
I A worthy goal: establish mechanistic explanations. But: ∗If this is upsetting, maybe string theory is for you... I For scientists, description is only part of the battle. I We still need to understand.
Frame 13/122 Frame 14/122
Overview of Overview of Super Basic definitions Complex Networks Super Basic definitions Complex Networks
Basic definitions Basic definitions
Examples of Examples of Nodes = A collection of entities which have Complex Networks Node degree = Number of links per node Complex Networks Properties of Properties of properties that are somehow related to each other Complex Networks Complex Networks I Notation: Node i’s degree = k . Nutshell i Nutshell I e.g., people, forks in rivers, proteins, webpages, Basic models of I ki = 0,1,2,. . . . Basic models of organisms,... complex networks complex networks Generalized random Notation: the average degree of a network = hki Generalized random networks I networks Scale-free networks (and sometimes z) Scale-free networks Small-world networks Small-world networks Links = Connections between nodes Generalized affiliation Generalized affiliation networks I Connection between number of edges m and networks References References I Links may be directed or undirected. average degree: 2m Links may be binary or weighted. hki = . I N
Other spiffing words: vertices and edges. I Defn: Ni = the set of i’s ki neighbors
Frame 15/122 Frame 16/122 Overview of Overview of Super Basic definitions Complex Networks Examples Complex Networks
Basic definitions Basic definitions
Examples of Examples of Adjacency matrix: Complex Networks Complex Networks Properties of Properties of I We represent a directed network by a matrix A with Complex Networks So what passes for a complex network? Complex Networks Nutshell Nutshell link weight aij for nodes i and j in entry (i, j). Basic models of I Complex networks are large (in node number) Basic models of I e.g., complex networks complex networks Generalized random I Complex networks are sparse (low edge to node Generalized random 0 1 1 1 0 networks networks Scale-free networks ratio) Scale-free networks 0 0 1 0 1 Small-world networks Small-world networks Generalized affiliation Generalized affiliation A = 1 0 0 0 0 networks I Complex networks are usually dynamic and evolving networks References References 0 1 0 0 1 I Complex networks can be social, economic, natural, 0 1 0 1 0 informational, abstract, ...
I (n.b., for numerical work, we always use sparse matrices.)
Frame 17/122 Frame 18/122
Overview of Overview of Examples Complex Networks Examples Complex Networks
Physical networks Basic definitions Interaction networks Basic definitions Examples of Examples of Complex Networks I The Blogosphere Complex Networks I River networks I The Internet Properties of Properties of Complex Networks I Biochemical Complex Networks I Neural networks I Road networks Nutshell networks Nutshell I Trees and leaves Power grids Basic models of Basic models of I complex networks I Gene-protein complex networks I Blood networks Generalized random Generalized random networks networks networks Scale-free networks Scale-free networks Small-world networks I Food webs: who Small-world networks Generalized affiliation Generalized affiliation networks eats whom networks References References I The World Wide Web (?)
I Airline networks
I Call networks I Distribution (branching) versus redistribution (AT&T) datamining.typepad.com () (cyclical) Frame 19/122 I The Media Frame 20/122 Overview of Overview of Examples Complex Networks Examples Complex Networks
Interaction networks: Basic definitions Basic definitions Examples of Examples of social networks Complex Networks Complex Networks Properties of Properties of I Snogging Complex Networks Complex Networks Nutshell Nutshell I Friendships Basic models of Basic models of Acquaintances complex networks complex networks I Generalized random Generalized random networks networks I Boards and Scale-free networks Scale-free networks Small-world networks Small-world networks directors Generalized affiliation Generalized affiliation networks networks I Organizations References References I facebook () twitter (), (Bearman et al., 2004)
I ‘Remotely sensed’ by: email activity, instant messaging, phone logs (*cough*).
Frame 21/122 Frame 22/122
Maps of Science
Results and Discussion Convergence. The FR algorithm can converge on different visualizations of the same network data. We do not claim Fig. 5 is According to the above mentioned methodology we constructed the only or best possible visualization. It was selected because it a map of science that visualizes the relationships between journals represents a particularly clear and uncluttered visualization of the according to user clickstreams. We first discuss the visual structure connections between journals in M’, and most importantly, its del.icio.us/url/830cfc21af4d02c392aa6ad04e93b93e 11/14/2006 09:04 PM of the map, and then attempt to validate the structural features of main structural features were stable across many different its underlying clickstream model by comparing the latter to journal iterations of the FR algorithm. centrality rankings and an alternative model of journal relations Connections. The journal connections shown in the map are derived from classification data. given by M’, not the FR algorithm. They are thus not artifacts of Overview of the visualization. Overview of A clickstream map of science Clustering. The FR algorithm will pull together small-scale Examples Complex Networks ClickworthyAny interpretation of the visual structure of Science: the map in Fig. 5 clusters of journals that are strongly connected in M’. The Complex Networks will be governed by the following considerations: appearance of small-scale journal clusters is thus directly related to Relational networks popular | recent del.icio.us / url Basic definitions Basic definitions login | register | help I Consumer purchases Examples of Examples of (Wal-Mart: ≈ 1 petabyte = 1015 bytes) Complex Networks Complex Networks Properties of Properties of Complex Networks Complex Networks » del.icio.us history for http://en.wikipedia.org/wiki/Main_PageI Thesauri: Networks of words generateddel.icio.us by meanings search Nutshell Nutshell I Knowledge/Databases/Ideas check url Basic models of Basic models of Metadata—Tagging: del.icio.us ( ) flickr ( ) complex networks complex networks I Generalized random Generalized random networks networks Scale-free networks Scale-free networks Small-world networks Small-world networks Main Page - Wikipedia, the free common tags cloud | list Generalized affiliation Generalized affiliation networks networks encyclopedia References References http://en.wikipedia.org/wiki/Main_Page community daily dictionary education encyclopedia this url has been saved by 16283 people. english free imported info information internet knowledge save this to your bookmarks » learning news reference research resource resources search tools useful web web2.0 wiki user notes wikipedia Figure 5. Map of science derived[5] from clickstream data. Circles represent individual journals. The lines that connect journals are the edges of theBollen clickstream model et in M’.al. Colors correspond to the AAT classification of the journal. Labels have been assigned to local clusters of journals that Nov ‘06 Frame 23/122 correspond to particular scientific disciplines. Frame 24/122 doi:10.1371/journal.pone.0004803.g005 all the knoweldge PLoS ONE | www.plosone.org 5 March 2009 | Volume 4 | Issue 3 | e4803 - Penfoldio related items - show ! facts - mmayes posting history I would not survive without Wikipedia, it knows » first posted by weblook to util everything, and I want to learn everything. - RobRemi Nov ‘06
The internet encyclopedia by faithfulsprig to reference - limannka by Penfoldio to wikipedia The mother of all wikis by vishinator to wikipedia - Username314 by mmayes to research Wikipedia - lancepickens by markcameron1 to system:unfiled by hirewad to system:unfiled ALWAYS useful! - AmritS by yeager.eng to firefox:toolbar Free Online Encyclopedia, User contributed articles by bandaids to reference - Maluvia by ingee to system:unfiled Welcome to Wikipedia, the free encyclopedia that by hdnev6 to system:unfiled anyone can edit. by hopabonk to encyclopedia - sanders_d by TianCaiWuDi to system:unfiled The Free Encyclopedia by k9entertainment to system:unfiled - puffseeker by llshoemaker to wikipedia Very good encycolopedia by tvn to imported kennis - slams by papaluukas to system:unfiled Knowledge by ureshi to system:unfiled - neilbriody by malph to reference encyclopedia Miscellaneous Info - sahing by salar15 to system:unfiled by RobRemi to wikipedia wiki knowledge encyclopedia free online encyclopedia http://del.icio.us/url/830cfc21af4d02c392aa6ad04e93b93e?settagview=cloud Page 1 of 65 Overview of Overview of A notable feature of large-scale networks: Complex Networks Properties Complex Networks
I Graphical renderings are often just a big mess. Basic definitions Basic definitions
Examples of Examples of Complex Networks Some key features of real complex networks: Complex Networks Properties of Properties of Complex Networks Complex Networks I Degree I Concurrency ⇐ Typical hairball Nutshell Nutshell distribution Hierarchical Basic models of I Basic models of I number of nodes N = 500 complex networks I Assortativity scaling complex networks Generalized random Generalized random I number of edges m = 1000 networks networks Scale-free networks I Homophily I Network distances Scale-free networks Small-world networks Small-world networks I average degree hki = 4 Generalized affiliation Clustering Generalized affiliation networks I I Centrality networks
References I Motifs I Efficiency References I And even when renderings somehow look good: I Modularity I Robustness “That is a very graphic analogy which aids understanding wonderfully while being, strictly I Coevolution of network structure speaking, wrong in every possible way” and processes on networks. said Ponder [Stibbons] —Making Money, T. Pratchett.
I We need to extract digestible, meaningful aspects. Frame 25/122 Frame 26/122
Overview of Overview of Properties Complex Networks Properties Complex Networks
Basic definitions Basic definitions
Examples of Examples of Complex Networks 2. Assortativity/3. Homophily: Complex Networks
1. Degree distribution Pk Properties of Properties of Complex Networks I Social networks: Homophily () = birds of a feather Complex Networks I Pk is the probability that a randomly selected node Nutshell Nutshell I e.g., degree is standard property for sorting: has degree k Basic models of Basic models of complex networks measure degree-degree correlations. complex networks Generalized random Generalized random I Big deal: Form of Pk key to network’s behavior networks [18] networks I Assortative network: similar degree nodes Scale-free networks Scale-free networks I ex 1: Erdös-Rényi random networks have a Poisson Small-world networks connecting to each other. Small-world networks Generalized affiliation Generalized affiliation distribution: networks networks I Often social: company directors, coauthors, actors. −hki k References References Pk = e hki /k! I Disassortative network: high degree nodes −γ I ex 2: “Scale-free” networks: Pk ∝ k ⇒ ‘hubs’ connecting to low degree nodes. We’ll come back to this business soon... I I Often techological or biological: Internet, protein interactions, neural networks, food webs.
Frame 27/122 Frame 28/122 Overview of Overview of Properties Complex Networks Properties Complex Networks
letter Basic definitions Basic definitions
4. Clustering: Examples of Examples of Complex Networks Network motifs in the transcriptional regulationComplex Networks I Your friends tend to know each other. Properties of network of Escherichia coli Properties of Complex Networks 5. Motifs: Complex Networks I Two measures: Nutshell Shai S. Shen-Orr1, Ron Milo2, Shmoolik Mangan1 & Uri Alon1,2 Nutshell I Small, recurring functional subnetworks *P + Basic models of Published online: 22 April 2002, DOI: 10.1038/ng881 Basic models of aj j complex networks complex networks j1j2∈Ni 1 2 [28] I e.g., Feed Forward Loop: Generalized random com Generalized random C = due to Watts & Strogatz . 1 networks networks 1–10 ( − )/ feedforward loop Little is known about the design principles of transcrip- ki ki 1 2 Scale-free networks a tional regulation networks thatScale-free contr networksol gene expression in i 2,11,12 Small-world networks X cells. Recent advances in dataSmall-world collection networks and analysis , X however, are generating unprecedented amounts of informa- Generalized affiliation Generalized affiliation networks tion about gene regulation networks.networks To understand these 1–10,13 3 × #triangles enetics.nature Y Y complex wiring diagrams , we sought to break down such [19] 2 C2 = due to Newman References networks into basic building blocksReferences. We generalize the notion Z n of motifs, widely used for sequence analysis, to the level of
#triples http://g networks. We define ‘network motifs’ as patterns of intercon-
nections that recur in many different parts of a network at fre- b crp oup [21] quencies much higher than those found in randomized
I C1 is the average fraction of pairs of neighbors who Shen-Orr, UriGr Alon, et al. networks. We applied new algorithms for systematically araC detecting network motifs to one of the best-characterized reg- ulation networks, that of direct transcriptional interactions in are connected. 3,6
lishing Escherichia coli . We find that much of the network is com- araBAD posed of repeated appearances of three highly significant Interpret C as probability two of a node’s friends Pub motifs. Each network motif has a specific function in determin- I 2 c single input module (SIM) ing gene expression, such as generating temporal expression programs and governing the responses to fluctuating external
know each other. Nature X X signals. The motif structure also allows an easily interpretable view of the entire known transcriptional network of the organ-
Frame 29/122 2002 ism. This approach may helpFrame define the 30/122 basic computational © n elements of other biological networks. Z1 Z2 ... Zn We compiled a data set of direct transcriptional interactions between transcription factors and the operons they regulate (an d operon is a group of contiguous genes that are transcribed into a single mRNA molecule). This database contains 577 interac- argR tions and 424 operons (involving 116 transcription factors); it was formed on the basis of on an existing database (Regu- lonDB)3,14. We enhanced RegulonDB by an extensive literature argI argF argD argE search, adding 35 new transcription factors, including alterna- argCBH tive σ-factors (subunits of RNA polymerase that confer recogni- tion of specific promoter sequences). The data set consists of Overview of e dense overlapping regulons (DOR) established interactions in which aOverview transcription of factor directly Properties Complex Networks Properties binds a regulatory site. Complex Networks The transcriptional network can be represented as a directed X1 X2 X3 ... Xn X1 X2 X3 Xn graph, in which each node represents an operon and edges repre- sent direct transcriptional interactions. Each edge is directed 6. modularity: Basic definitions Basic definitions Z1 Z2 Z3 Z4 ... Zm Fig. 1 Network motifs found in the E. coli transcriptional regulation network. Examples of Symbols representing the motifs areExamples also shown. a, Feedforward of loop: a tran- scription factor X regulates a second transcription factor Y, and both jointly Complex Networks f regulate one or more operons Z1...ZnComplex. b, Example of Networksa feedforward loop (L-ara- 49 binose utilization). c, SIM motif: a single transcription factor, X, regulates a set crp 53 rpoS ada oxyR ihf lrp hns rcsA nhaR fis 58 of operons Z1...Zn. X is usually autoregulatory. All regulations are of the same 63 46 Properties of sign. No other transcription factor regulatesProperties the operons. of d, Example of a SIM 83 114 system (arginine biosynthesis). e, DOR motif: a set of operons Z1...Zm are each 33 28 Complex Networks Complex Networks 25 11 7. Concurrency: regulated by a combination of a set of input transcription factors, X1...Xn. 97 88 dps DORs are defined by an algorithm that detects dense regions of connections, 1 59 alkA katG proP 67 nhaA osmC with a high ratio of connections to transcription factors. f, Example of a DOR 73 ftsQAZ 105 24 Nutshell Nutshell 50 (stationary phase response). 103 37 89 69 36 Transmission of a contagious element only occurs 45 109 110 Basic models of I Basic models of 57 90 complex networks [16]1Department of Molecular Cell Biology, 2Department of Physics of Complex Systems, Weizmann Institute of Science,complex Rehovot 76100, networks Israel. Correspondence 44 66 34 should be addressed to U.A. (e-mail: [email protected]). 42 Generalized random during contact Generalized random 16 82 4 75 31 networks networks 86 93 91 112 80 0 Scale-free networks 64 naturScale-freee genetics networks • volume 31 • may 2002 48 18 54 Rather obvious but easily missed in a simple model 9 92 Small-world networks I Small-world networks
23 7 104 29 8 61 Generalized affiliation Generalized affiliation 94 71 41 78 35 networks networks 68 I Dynamic property—static networks are not enough 99 22 19 55 21 77 References References 5 10 111 81 101 30 I Knowledge of previous contacts crucial 3 79 108 51 85 38 52 84 Beware cumulated network data! 98 I 2 6 113 17 43 26 76 70 107 60 39 40 14 74 72 47 62 13 95 27 96 12 100 15 102 65 20 87 106 64 32 56 Clauset et al., 2006 [8]: NCAA football Frame 31/122 Frame 32/122 Overview of Overview of Properties Complex Networks Properties Complex Networks
8. Horton-Strahler stream ordering: Basic definitions Basic definitions
Examples of Examples of I Metrics for branching networks: Complex Networks 9. Network distances: Complex Networks I Method for ordering streams hierarchically Properties of Properties of Complex Networks (a) shortest path length d : Complex Networks I Reveals fractal nature of natural branching networks ij [23, 10] Nutshell Nutshell I Hierarchy is not pure but mixed (Tokunaga). Basic models of I Fewest number of steps between nodes i and j. Basic models of I Major examples: rivers and blood networks. complex networks complex networks Generalized random Generalized random networks I (Also called the chemical distance between i and j.) networks Scale-free networks Scale-free networks Small-world networks Small-world networks Generalized affiliation Generalized affiliation networks (b) average path length hdij i: networks References References I Average shortest path length in whole network.
I Good algorithms exist for calculation.
I Weighted links can be accommodated.