Control Issues in Cancer Therapy

在这项建议中概述的ESEA的目标有两个方面：第一，我们打算探索应用现有的理论知识来推进癌症治疗事业的可能性；第二，我们打算发展控制理论的基本结果，以解决在癌症相关研究中出现的一些独特的问题。提出的研究是由P.I.S在国家癌症研究所资助的两年生物信息学教授期间的教学和研究经验推动的。在这次培训的过程中，P.I.已经认识到癌症是由控制系统的崩溃引起的，该系统控制了多细胞组织中的细胞数量S，并阻止了像S这样的不良事件的发生P.I.和共同P.I.S坚信，工程文献中经过时间考验的控制技术可以得到应用，并进一步发展，以获得癌症治疗的有用指南。已经获得了明确方向的省略结果。具体地说，在与我们的分子生物学家同事Michael Bittner(与国家人类基因组研究所合作，目前与T翻译基因组研究所合作)的合作下，我们发展了一种建模受外部干预影响的遗传调控网络的工作，并用它来应用最优控制策略来减缓转移性黑色素瘤的进展。所用的建模方法是带有外部控制的P-obbilistic Boolean Network(PBN)模型，我们证明了这是受控马尔可夫链的一个特例。这使得应用动态P-amaming等最优控制技术成为可能，并使我们能够进行这一概念的验证。受这一成功和我们在控制论方面的丰富经验的启发，我们已经开始了一项扩展的研究，旨在模拟癌细胞受到不同剂量电离辐射时的行为。这项研究是与国家癌症研究所的阿尔伯特·福纳斯的S实验室合作进行的，这是我们的主要数据来源。利用这些放射研究的基因实验数据，我们打算建立PBN模型，我们希望这将使我们能够展望未来，当细胞死亡时，它们被称为p53的基因。这是一个容易的控制目标，也具有生物学意义，因为(I)受到辐射的细胞遭受dna损伤；(Ii)如果允许DNA损伤的细胞复制，可能会导致癌症；和(Iii)P53基因的功能之一是确保细胞分裂的DNA复制阶段不发生，除非亲本细胞的DNA损伤被证实。我们对辐射实验的基因表达数据的初步评估在具有功能的P53的细胞系和具有缺失的P53的细胞系之间表现出显著的差异。我们计划追求的ESEA路线S是一个非常完美的人。这个提议的一个协作者S在将信号处理技术从工程学文献应用到基因组学领域方面起到了重要作用，而另一个协作者对对照领域有很长的原始贡献。对拟议研究成功的乐观部分源于这样一个事实，即P.I.已经成功地与两个协作者都进行了ESEA合作。此外，S在过去让我们的分子生物学家S、迈克尔·比特纳和阿尔伯特·福恩斯充分参与ESEA，将大大有助于确保研究结果具有生物学意义。我们在癌症治疗方面取得的任何进展，即使是在细胞系方面，都必将对社会产生深远的影响。使用细胞系进行成功的干预研究是导致后续动物研究和最终临床试验的先导，在癌症治疗被批准用于普通人群之前。因此，任何有意义的理由都表明，我们在这项研究的基础上取得的进展不仅会影响目前的癌症治疗但也是在未来的几年里。这种好处将是拯救下来的人类生命的一部分

The objectives of the esea ch outlined in this proposal a e two fold.First,we intend to explo e thepossibility of applying existing cont ol theo etic knowledge to advance the cause of cancer therapy;andsecond,we intend to develop fundamental esults in control theo y to address some of the unique p oblemsthat arise in cancer related esearch.The proposed esearch is motivated by the P.I.'s didactic and esearchexperiences du ing his two-year stint as a Bioinformatics T ainee supported by a National Cancer InstituteT aining Grant.In the course of this training,the P.I.has ealized that the disease Cancer is causedby a b eakdown in the control system that egulates the cell numbe s in a multicellula o ganismand p events undesirable g owths such as tumo s from developing.Furthermo e,the P.I.and the co-P.I.'s strongly believe that the time-tested cont ol techniques from the engineering literature could beha nessed,and further developed fo obtaining useful guidelines in cancer therapy.P omising p elimina y esults indicating a clea esea ch direction have already been obtained.Specif-ically,in collabo ation with ou molecula biologist colleague,Michael Bittner (fo merly with the Na-tional Human Genome Resea ch Institute (NHGRI),cur ently with the T anslational Genomics Institute(TGEN)),we developed a f amework for modeling genetic regulatory networks subject to external in-terventions and used it to apply optimal control st ategies to slow down the p ogression of metastaticmelanoma.The pa ticula modeling f amewo k used was the P obabilistic Boolean Netwo k (PBN)modelwith external controls,which we showed to be a special case of a Cont olled Markov Chain.This madeit possible to apply standa d optimal control techniques such as Dynamic P og amming and enabled usto carry out this proof of concept study.Motivated by this success and our extensive experience in Control Theory,we have embarked on anexpanded study aimed at modeling the behaviour of cancer cell lines when they are subjected to variousdoses of ionizing adiation.This study is being carried out in collaboration with Albert Fornace 's labo-ato y at the National Cancer Institute (NCI),which is our prima y data sou ce.Using gene exp essiondata from these adiation studies,we intend to build PBN models that we hope will enable us to look foways to increase the capacity of cells to die upon adiation when they are de .cient in a gene called p53.This is a easonable control objective that also makes biological sense since (i)cells subjected to radiationundergo DNA damage;(ii)if the cell with DNA damage is allowed to eplicate,cancer can esult;and(iii)one of the functions of the gene p53 is to ensure that the DNA eplication stage of cell division doesnot occu unless the damaged DNA of the parent cell is .rst epaired.Ou initial assessment of geneexpression data f om radiation experiments has evealed marked di .erences between the cell lines withfunctional p53 versus those with de .cient p53.Consequently,the line of esea ch that we a e planningto pursue appea s to be a very p omising one.One of the co-P.I.'s of this proposal was instrumental inpioneering the application of signal p ocessing techniques from the engineering literature to the .eld ofgenomics while the other co-P.I.has had a long histo y of o iginal contributions to the Controls .eld.Optimism about the success of the proposed research partially stems from the fact that the P.I.has hadsuccessful esea ch collabo ations with both the co-P.I.'s in the past.In addition,keeping our moleculabiologist collabo ato s Michael Bittner and Albert Fo nace fully involved in the esea ch will go a longway towards ensuring that the esults developed emain biologically meaningful.Any advance that we make in pursuing cancer therapy,even fo cell lines,is bound to have a profoundimpact on society.Successful intervention studies using cell lines are the precursors that lead to subsequentanimal studies and ultimately clinical trials,befo e the pa ticula cancer treatment is approved fo useon the general population.Thus,any signi .cant ground that we b eak in the cou se of this research willimpact cancer treatment not only fo the present time,but also fo years to come.The bene .t will be interms of human lives that are saved down the oad.2