git-svn-id: file:///home/git/hedgehog.fhcrc.org/bioconductor/trunk/madman/Rpacks/Rdisop@64639 bc3139a8-67e5-0310-9ffc-ced21a209358
Steffen Neumann authored on 30/03/2012 05:51:40
| 1 | 1 |
new file mode 100644 |
| ... | ... |
@@ -0,0 +1,149 @@ |
| 1 |
+@Article{boecker08decomp,
|
|
| 2 |
+author = {Sebastian B\"ocker and Zsuzsanna Lipt\'ak and Marcel Martin and Anton Pervukhin and Henner Sudek},
|
|
| 3 |
+title = {{DECOMP}---from interpreting Mass Spectrometry peaks to solving the {Money} {Changing} {Problem}},
|
|
| 4 |
+journal = {Bioinformatics},
|
|
| 5 |
+year = {2008},
|
|
| 6 |
+volume = {24},
|
|
| 7 |
+number = {4},
|
|
| 8 |
+pages = {591-593},
|
|
| 9 |
+url = {http://bioinformatics.oxfordjournals.org/cgi/reprint/24/4/591?ijkey=1lM50Bkzz4SCLsa&keytype=ref},
|
|
| 10 |
+doi = {10.1093/bioinformatics/btm631},
|
|
| 11 |
+abstract = {We introduce DECOMP, a tool that computes the sum formula
|
|
| 12 |
+ of all molecules whose mass equals the input |
|
| 13 |
+ mass. This problem arises frequently in biochemistry |
|
| 14 |
+ and mass spectrometry (MS), when we know the |
|
| 15 |
+ molecular mass of a protein, DNA, or metabolite |
|
| 16 |
+ fragment but have no other information. A closely |
|
| 17 |
+ related problem is known as the Money Changing |
|
| 18 |
+ Problem (MCP), where all masses are positive |
|
| 19 |
+ integers. Recently, efficient algorithms have been |
|
| 20 |
+ developed for the MCP, which DECOMP applies to real |
|
| 21 |
+ valued MS data. The excellent performance of this |
|
| 22 |
+ method on proteomic and metabolomic MS data has |
|
| 23 |
+ recently been demonstrated. DECOMP has an |
|
| 24 |
+ easy-to-use graphical interface, which caters for |
|
| 25 |
+ both types of users: those interested in solving MCP |
|
| 26 |
+ instances and those submitting MS data.}, |
|
| 27 |
+} |
|
| 28 |
+ |
|
| 29 |
+@Article{boecker09sirius,
|
|
| 30 |
+author = {Sebastian B\"ocker and Matthias Letzel and Zsuzsanna Lipt{\'a}k and Anton Pervukhin},
|
|
| 31 |
+title = {{SIRIUS}: Decomposing isotope patterns for metabolite identification},
|
|
| 32 |
+journal = {Bioinformatics},
|
|
| 33 |
+year = {2009},
|
|
| 34 |
+volume = {25},
|
|
| 35 |
+number = {2},
|
|
| 36 |
+pages = {218--224},
|
|
| 37 |
+url = {http://bioinformatics.oxfordjournals.org/cgi/content/full/25/2/218},
|
|
| 38 |
+doi = {10.1093/bioinformatics/btn603},
|
|
| 39 |
+abstract = {Motivation: High-resolution mass spectrometry (MS) is
|
|
| 40 |
+ among the most widely used technologies in |
|
| 41 |
+ metabolomics. Metabolites participate in almost all |
|
| 42 |
+ cellular processes, but most metabolites still |
|
| 43 |
+ remain uncharacterized. Determination of the sum |
|
| 44 |
+ formula is a crucial step in the identification of |
|
| 45 |
+ an unknown metabolite, as it reduces its possible |
|
| 46 |
+ structures to a hopefully manageable set. Results: |
|
| 47 |
+ We present a method for determining the sum formula |
|
| 48 |
+ of a metabolite solely from its mass and the natural |
|
| 49 |
+ distribution of its isotopes. Our input is a |
|
| 50 |
+ measured isotope pattern from a high resolution mass |
|
| 51 |
+ spectrometer, and we want to find those molecules |
|
| 52 |
+ that best match this pattern. Our method is |
|
| 53 |
+ computationally efficient, and results on |
|
| 54 |
+ experimental data are very promising: For orthogonal |
|
| 55 |
+ time-of-flight mass spectrometry, we correctly |
|
| 56 |
+ identify sum formulas for more than 90\% of the |
|
| 57 |
+ molecules, ranging in mass up to 1000 |
|
| 58 |
+ Da. Availability: SIRIUS is available under the LGPL |
|
| 59 |
+ license at |
|
| 60 |
+ http://bio.informatik.uni-jena.de/sirius/. Contact: |
|
| 61 |
+ anton.pervukhin@minet.uni-jena.de}, |
|
| 62 |
+} |
|
| 63 |
+ |
|
| 64 |
+@Inproceedings{boecker06decomposing,
|
|
| 65 |
+author = {Sebastian B\"ocker and Matthias Letzel and Zsuzsanna Lipt{\'a}k and Anton Pervukhin},
|
|
| 66 |
+title = {Decomposing metabolomic isotope patterns},
|
|
| 67 |
+booktitle = {Proc. of Workshop on Algorithms in Bioinformatics (WABI 2006)},
|
|
| 68 |
+publisher = {Springer, Berlin},
|
|
| 69 |
+year = {2006},
|
|
| 70 |
+volume = {4175},
|
|
| 71 |
+pages = {12--23},
|
|
| 72 |
+series = {Lect. Notes Comput. Sci.},
|
|
| 73 |
+url = {http://bio.informatik.uni-jena.de/bib2html/downloads/2006/BoeckerEtAl_DecomposingMetabolomicIsotopePatterns_WABI_2006.pdf},
|
|
| 74 |
+abstract = {We present a method for determining the sum formula of
|
|
| 75 |
+ metabolites solely from their mass and isotope |
|
| 76 |
+ pattern. Metabolites, such as sugars or lipids, |
|
| 77 |
+ participate in almost all cellular processes, but |
|
| 78 |
+ the majority still remains uncharacterized. Our |
|
| 79 |
+ input is a measured isotope pattern from a high |
|
| 80 |
+ resolution mass spectrometer, and we want to find |
|
| 81 |
+ those molecules that best match this |
|
| 82 |
+ pattern. Determination of the sum formula is a |
|
| 83 |
+ crucial step in the identification of an unknown |
|
| 84 |
+ metabolite, as it reduces its possible structures to |
|
| 85 |
+ a hopefully manageable set. Our method is |
|
| 86 |
+ computationally efficient, and first results on |
|
| 87 |
+ experimental data indicate good identification rates |
|
| 88 |
+ for chemical compounds up to 700 Dalton. Above 1000 |
|
| 89 |
+ Dalton, the number of molecules with a certain mass |
|
| 90 |
+ increases rapidly. To efficiently analyze mass |
|
| 91 |
+ spectra of such molecules, we define several |
|
| 92 |
+ additive invariants extracted from the input and |
|
| 93 |
+ then propose to solve a joint decomposition |
|
| 94 |
+ problem.}, |
|
| 95 |
+} |
|
| 96 |
+ |
|
| 97 |
+@Article{boecker07fast,
|
|
| 98 |
+author = {Sebastian B\"ocker and Zsuzsanna Lipt{\'a}k},
|
|
| 99 |
+title = {A fast and simple algorithm for the {M}oney {C}hanging {P}roblem},
|
|
| 100 |
+journal = {Algorithmica},
|
|
| 101 |
+year = {2007},
|
|
| 102 |
+volume = {48},
|
|
| 103 |
+number = {4},
|
|
| 104 |
+pages = {413-432},
|
|
| 105 |
+doi = {10.1007/s00453-007-0162-8},
|
|
| 106 |
+abstract = {The Money Changing Problem (MCP) can be stated as follows:
|
|
| 107 |
+ Given $k$ positive integers $a_1< \ldots < a_k$ and |
|
| 108 |
+ a query integer $M$, is there a linear combination |
|
| 109 |
+ $\sum_i c_ia_i = M$ with non-negative integers |
|
| 110 |
+ $c_i$, a \emph{decomposition} of $M$? If so, produce
|
|
| 111 |
+ one or all such decompositions. The largest integer |
|
| 112 |
+ without such a decomposition is called the |
|
| 113 |
+ \emph{Frobenius number} $g(a_1,\ldots,a_k)$. A data
|
|
| 114 |
+ structure called {\em residue table} of $a_1$ words
|
|
| 115 |
+ can be used to compute the Frobenius number in time |
|
| 116 |
+ $O(a_1)$. We present an intriguingly simple |
|
| 117 |
+ algorithm for computing the residue table which runs |
|
| 118 |
+ in time $O(ka_1)$, with no additional memory |
|
| 119 |
+ requirements, outperforming the best previously |
|
| 120 |
+ known algorithm. Simulations show that it performs |
|
| 121 |
+ well even on 'hard' instances from the |
|
| 122 |
+ literature. In addition, we can employ the residue |
|
| 123 |
+ table to answer MCP decision instances in constant |
|
| 124 |
+ time, and a slight modification of size $O(a_1)$ to |
|
| 125 |
+ compute one decomposition for a query $M$. Note that |
|
| 126 |
+ since both, computing the Frobenius number and MCP |
|
| 127 |
+ (decision) are NP-hard, one cannot expect to find an |
|
| 128 |
+ algorithm that is polynomial in the size of the |
|
| 129 |
+ input, i.e., in $k,\log a_k$, and $\log M$. We then |
|
| 130 |
+ give an algorithm which, using a modification of the |
|
| 131 |
+ residue table, also constructible in $O(ka_1)$ time, |
|
| 132 |
+ computes all decompositions of a query integer |
|
| 133 |
+ $M$. Its worst-case running time is $O(ka_1)$ for |
|
| 134 |
+ each decomposition, thus the total runtime depends |
|
| 135 |
+ only on the output size and is independent of the |
|
| 136 |
+ size of query $M$ itself. We apply our latter |
|
| 137 |
+ algorithm to interpreting mass spectrometry (MS) |
|
| 138 |
+ peaks: Due to its high speed and accuracy, MS is now |
|
| 139 |
+ the method of choice in protein |
|
| 140 |
+ identification. Interpreting individual peaks is one |
|
| 141 |
+ of the recurring subproblems in analyzing MS data; |
|
| 142 |
+ the task is to identify sample molecules whose mass |
|
| 143 |
+ the peak possibly represents. This can be stated as |
|
| 144 |
+ an MCP instance, with the masses of the individual |
|
| 145 |
+ amino acids as the $k$ integers $a_1,\ldots, |
|
| 146 |
+ a_k$. Our simulations show that our algorithm is |
|
| 147 |
+ fast on real data and is well suited for generating |
|
| 148 |
+ candidates for peak interpretation.}, |
|
| 149 |
+} |
|
| 0 | 150 |
\ No newline at end of file |