1 The field of Genome annotation pay a lot of attentions where the ability to collect and analysis genomical data can provide strong indicator for the study of life\cite{Eisen2007}. A lot of genome annotation centres present various types of annotations tools (i.e cost-effective sequencing methods\cite{Bakke2009}) on different annotation levels. Two method of gene finding in annotated genome can be categorized as: Alignment-based, composition based, or combination of both\cite{parra2007cegma}. The Alignment-based method is used when we try to predict a coding gene (i.e. Genes that produce proteins) by aligning DNA sequence of gene to the protein of cDNA sequence of homolog\cite{parra2007cegma}. This approache also used in GeneWise\cite{birney2004genewise} with known splicing signals. Composition-based mothod (known as \textit{ab initio} is based on a probabilistic model of gene structure to find genes and/or new genes accoding to the probability gene value, this method like GeneID\cite{parra2000geneid}. In this section, we will consider a new method of finding core genes from large amount of chloroplast genomes, as a solution of the previous method where stated in section two. This method is based on extracting gene features. The question now is how can we have good annotation genome? To answer this question, we need to focusing on studying the annotation accuracy\cite{Bakke2009}) of the genome. A general overview of the system is illustrated in Figure \ref{Fig1}.\\
5 \includegraphics[width=0.7\textwidth]{generalView}
6 \caption{A general overview of the system}\label{Fig1}
9 In Figure 1, we illustrate the general overview of system pipeline: \textit{Database, Genomes annotation, Gene extraction, } and \textit{relationships}. We will give a short discussion for each stage in the model in order to understand all core extraction process. Good database (as a first stage) will produce good results, however, many international Banks for nucleotide sequence databases like (GenBank in USA, EMBL-Bank in Europe, and DDBJ in Japon) where exists to store various genomes and DNA species. A lot of Biological tool interact with these databases for (Genome Annotation, Gene extraction, alignments, ... , etc). The database in model must be any confident data source that store annotated and/or unannotated chloroplast genomes. We will consider GenBank- NCBI database to be our nucleotide sequences database. Annotation (as the second stage) is consider to be the first important task for Extract Gene Features. Thanks to good annotation tool that lead us to extract good gene features. In this paper, two annotation techniques from \textit{NCBI, and Dogma} will be used to extract \textit{genes features}. Extracting Gene feature (as a third stage) can be anything like (genes names, gene sequences, protein sequence,...etc). Our methodologies in this paper will consider gene names, gene counts, and gene sequences for extracting core genes and chloroplast evolutionary tree. \\
10 In last stage, verifying the work from Biological expert needs to organize and represent genomes relationships and gene evolution in the form of (tables, phylogenetic trees, graphs,...,etc). In addition, comparing these forms with the results from another annotation tool like Dogma\cite{RDogma} for large population of chloroplast genomes give to us biological perspective to the nature of chloroplast evolution. \\
11 A Local database attache with each pipe stage to store all information of extraction process. The output from each stage in our system will be an input to the second stage and so on.
13 \subsection{Genomes Samples}
14 In this research, we retrieved 107 genomes of Chloroplasts from NCBI where 8 genomes considered to be not good. The remain 99 genomes lies in the 11 types of chloroplast families, as shown in Table \ref{Tab1}. The list of distribution of genomes is illstrated in detail in Table \ref{Tab2}.
17 \caption{distribution on Chloroplast Families}\label{Tab1}
21 Family & Genome Counts \\ [0.5ex]
28 Dinoflagellates & 02 \\
32 Lycopodiophyta & 02 \\
33 Haptophytes & 01 \\ [1ex]
38 \input{population_Table}
40 \subsection{Genome Annotation Techniques}
41 The second stage in system pipeline is genome annotation. Many annotation techniques were developed for annotate chloroplast genomes but they vary in the number and type of predicting genes (i.e the ability to predict genes and \textit{Transfere RNA (tRNA)} and \textit{Ribosomal RNA (rRNA)} genes). Two annotation techniques from NCBI and Dogma are considered to analyse chloroplast genomes to examin the accuricy of predicted coding genes. Figure \ref{NCBI_annotation}, illstrate two annotation technique.\\
45 \includegraphics[width=0.7\textwidth]{NCBI_annotation}
46 \caption{Genome annotation using either NCBI or Dogma}\label{NCBI_annotation}
49 With each annotation model, we provide a quality check class for the flow of chloroplast genomes. This class has an access to NCBI taxonomy database based on genome accession number to retreive information for the genome. These information contains \textit{[Scientific name, lineage, Division, taxonomy ID, parentID, and Accession No]}. Examin each genome with this class (i.e based on some parameters), can ignore some genomes from this competition that not match a specific control condition.
51 \subsubsection{genome annotation from NCBI}
52 The objective from this step is to organize, solve genes duplications, and generate sets of genes from each genome. The input to the system is our list of chloroplast genomes, annotated from NCBI\cite{Sayers01012011}. All genomes stored as \textit{.fasta} files include collection of Protein coding genes\cite{parra2007cegma,RDogma}(gene that produce proteins) with its coding sequences.
53 As a preparation step to achieve the set of core genes, we need to translate these genomes using \textit{BioPython} package\cite{chapman2000biopython}, and extracting all information needed to find the core genes. A process starts by converting each genome in fasta format to GenVision\cite{geneVision} formats from DNASTAR, and this is not an easy job. The output from this operation is a lists of genes stored in a local database for genomes, their genes names and gene counts. In this stage, we will accumulate some Gene duplications with each genome treated. In other words, duplication in gene name can comes from genes fragments as long as chloroplast DNA sequences. We defines \textit{Identical state} to be the state that each gene present only one time in a genome (i.e Gene has no copy) without considering the position or gene orientation. This state can be reached by filtering the database from redundant gene name. To do this, we have two solutions: first, we made an orthography checking. Orthographe checking is used to merge fragments of a gene to form one gene.
54 Second, we convert the list of genes names for each genome (i.e. after orthography check) in the database to be a set of genes names. Mathematically speaking, if $G=\left[g_1,g_2,g_3,g_1,g_3,g_4\right]$ is a list of genes names, by using the definition of a set in mathematics, we will have $set(G)=\{g_1,g_2,g_3,g_4\}$, and $|G|=4$ where $|G|$ is the cardinality number of the set $G$ which represent the number of genes in the set.\\
55 The whole process of extracting core genome based on genes names and counts among genomes is illustrate in Figure \ref{NCBI:Annotation}.\\
59 \includegraphics[width=0.7\textwidth]{NCBI_GeneName}
60 \caption{NCBI Annotation for Chloroplast genomes}
61 \label{NCBI:Annotation}
64 \subsubsection{Genome annotation from Dogma}
65 Dogma is an annotation tool developed in the university of Texas by \cite{RDogma} in 2004. Dogma is an abbreviation of \textit{Dual Organellar GenoMe Annotator}\cite{RDogma} for plant chloroplast and animal mitochondrial genomes.
66 It has its own database for translated the genome in all six reading frames and query the amino acid sequence database using Blast\cite{altschul1990basic}(i.e Blastx) with various parameters, and to identify protein coding genes\cite{parra2007cegma,RDogma} in the input genome based on sequence similarity of genes in Dogma database. Further more, it can produce the \textit{Transfer RNAs (tRNA)}\cite{RDogma}, and the \textit{Ribosomal RNAs (rRNA)}\cite{RDogma} and verifying their start and end positions rather than NCBI annotation tool. There are no gene duplication with dogma after solving gene fragmentation. \\
67 Genome Anntation with dogma can be the key difference of extracting core genes. In figure \ref{dog:Annotation}, The step of annotation divided into two tasks: First, It starts to annotate complete choloroplast genomes (i.e \textit{Unannotated genomes} from NCBI by using Dogma web tool. The whole annotation process was done manually. The output from dogma is considered to be collection of coding genes file for each genome in the form of GeneVision\cite{geneVision} file format.\\
68 Where the second task is to solve gene fragments. Defragment process starts immediately after the first task to solve fragments of coding genes for each genome to avoid gene duplication. All genomes after this stage are fully annotated, their genes were de-fragmented, genes lists and counts were identified. These information stored in local database.\\
71 \includegraphics[width=0.7\textwidth]{Dogma_GeneName}
72 \caption{Dogma Annotation for Chloroplast genomes}\label{dog:Annotation}
75 From these two tasks, we can obtain clearly one copy of coding genes. To ensure that genes produced from dogma annotation process is same as the genes in NCBI. We apply in parrallel a quality checking process that align each gene from dogma and NCBI with respect to a specific threshold.\\
77 \subsection{Extract Gene Features}
78 The goal of this step is trying to find maximum core genes from sets of genes (\textit{Vectors}) where stored in the local database from the annotation process. The key of finding core genes is to collect from each iteration of genes comparisons the maximum number of common genes. To do so, the system build an \textit{Intersection core matrix(ICM)}. ICM here is a two dimensional symmetric matrix (considered as a vector space) where each row and column represent a vector for one genome. Each position in ICM stores the \textit{intersection scores}. Intersection Score(IS) is the cardinality number of a core genes comes from intersecting one vector with other vectors in vector space. Taking maximum cardinality from each row and then take the maximum of them will result to select the best two genomes with their maximum core. Mathematically speaking, if we have an $m \times n$ vector space matrix where $m=n=$number of vectors in local database, then lets consider:\\
81 Score=\max_{i<j}\vert x_i \cap x_j\vert
85 Where $x_i, x_j$ are vectors. To constract new core genes, we look on the value of intersecting two vectors, we call it $Score$:\\
86 $$\text{New Core} = \begin{cases}
87 \text{Ignored} & \text{if $Score=0$;} \\
88 \text{new Core id} & \text{if $Score>0$.}
91 if $Score=0$ then we have \textit{disjoint relation} (i.e no common genes between two genomes). In this case the system ignore the vector that smash the core genes. Otherwise, The system remove these two vectors from ICM and add new core vector with a \textit{coreID} of them to ICM for the calculation in next iteration. The partial core vectors generated with its values will store in the local database for reused to draw the tree. this process repeat until all vectors treated.
92 We observe that ICM will result to be very large because of the huge amount of data that it stores. In addition, this will results to be time and memory consuming for calculating the intersection scores by using just genes names. To increase the speed of calculations, we can calculate the upper triangle scores only and exclude diagonal scores. This will reduce whole processing time and memory to half. The time complexity for this process after enhancement changed from $O(n^2-n)$ to $O(\frac{(n-1).n}{2})$.\\
94 The Algorithm of construction the vector matrix and extracting the vector of maximum core genes where illustrated in Algorithm \ref{Alg1}. The output from this step is the maximum core vector with its two vectors to draw it in a tree.
97 \caption{Extract Maximum Intersection Score}
100 \REQUIRE $L \leftarrow \text{genomes vectors}$
101 \ENSURE $B1 \leftarrow Max core vector$
102 \FOR{$i \leftarrow 0:len(L)-1$}
103 \STATE $core1 \leftarrow set(GenomeList[L[i]])$
104 \STATE $score1 \leftarrow 0$
105 \STATE $g1,g2 \leftarrow$ " "
106 \FOR{$j \leftarrow i+1:len(L)$}
107 \STATE $core2 \leftarrow set(GenomeList[L[i]])$
109 \STATE $Core \leftarrow core1 \cap core2$
110 \IF{$len(Core) > score1$}
111 \STATE $g1 \leftarrow L[i]$
112 \STATE $g2 \leftarrow L[j]$
113 \STATE $Score \leftarrow len(Core)$
114 \ELSIF{$len(Core) == 0$}
115 \STATE $g1 \leftarrow L[i]$
116 \STATE $g2 \leftarrow L[j]$
117 \STATE $Score \leftarrow -1$
121 \STATE $B1[score1] \leftarrow (g1,g2)$
127 \textit{GenomeList} represents the local database.\\
129 \subsection{Genomes Relationships}
130 The goal here is to visualizing the results by build a tree of evolution. The system can produce this tree automatically by using Dot graphs package\cite{gansner2002drawing} from Graphviz library and all information available in a database. Core genes generated with their genes can be very important information in the tree, because they can viewed as an ancestor information for two genomes or more. Further more, each node represents a genome or core as \textit{(Genes count:Family name, Scientific names, Accession number)}, Edges represent numbers of lost genes from genomes-core or core-core relationship. The number of lost genes here can represent an important factor for evolution, it represents how much lost of genes for the species in same or different families. By the principle of classification, small number of gene lost among species indicate that those species are close to each other and belong to same family, while big genes lost means that species is far to be in the same family. To see the picture clearly, Phylogenetic tree is an evolutionary tree generated also by the system. Generating this tree is based on the distances among genes sequences. There are many resources to build such tree (for example: PHYML\cite{guindon2005phyml}, RAxML{\cite{stamatakis2008raxml,stamatakis2005raxml}, BioNJ , and TNT\cite{goloboff2008tnt}}. We consider to use RAxML\cite{stamatakis2008raxml,stamatakis2005raxml} to generate this tree.
132 \section{Implementation}
133 We implemented four algorithms to extract maximum core genes from large amount of chloroplast genomes. Two algorithms used to extract core genes based on NCBI annotation, and the others based on dogma annotation tool. Evolutionary tree generated as a result from each method implementation. In this section, we will present the four methods, and how they can extract maximum core genes?, and how the developed code will generate the evolutionary tree.
135 \subsection{Extract Core genes based on NCBI Gene names and counts}
136 The first idea to construct the core genome is based on the extraction of Genes names (as gene presence or absence). For instant, in this stage neither sequence comparison nor new annotation were made, we just want to extract all genes with counts stored in each chloroplast genome, then find the intersection core genes based on gene names. \\
137 The pipeline of extracting core genes can summarize in the following steps:\\
138 First, we apply the genome annotation method using NCBI annotation tool. Genome quality check can be used in this step to ensure that genomes pass some quality condition. Then, the system lunch annotation process using NCBI to extract code genes (i.e \textit{exons}) and solve gene fragments. From NCBI, we did not observe any problem with genes fragments, but there are a problem of genes orthography (e.g two different genes sequences with same gene name). After we obtain all annotated genomes from NCBI to the local database, the code will then automatically will generate GenVision\cite{geneVision} file format to lunch the second step to extract coding genes names and counts. The competition will start by building intersection matrix to intersect genomes vectors in the local database with the others. New core vector for two leaf vectors will generate and a specific \textit{CoreId} will assign to it. an evolutionary tree will take place by using all data generated from step 1 and 2. The tree will also display the amount of genes lost from each intersection iteration. A specific excel file will be generated that store all the data in local database. The whole operation illstrate in Figure \ref{NCBI:geneextraction}.
142 \includegraphics[width=0.7\textwidth]{NCBI_geneextraction}
143 \caption{Extract core genes based on NCBI gene names and counts}\label{NCBI:geneextraction}
146 \subsection{Extract Core genes based on Dogma Gene names and counts}
147 The main goal is to get as much as possible the core genes of maximum coding genes names. According to NCBI annotation problem based on \cite{Bakke2009}, annotation method like dogma can give us more reliable coding genes than NCBI. This is because NCBI annotation can carry some annotation and gene identification errors. The general overview of whole process of extraction illustrated in figure \ref{dog:Annotation}. From this figure, the pipeline of extracting core genes can summarize in the following steps:\\
148 First, we apply the genome annotation method using Dogma annotation tool. Genome quality check can be used in this step to ensure that genomes pass some quality condition. Then, the system lunch annotation process using Dogma to extract code genes (i.e \textit{exons}) and solve gene fragments. The key difference here is that dogma can generate in addition transfer RNA and ribosomal RNA. As a result from annotation process with dogma is genomes files in GenVision\cite{geneVision} file format, the code will lunch genes de-fragments process to avoid genes duplications. little problems of genes orthography (e.g two different genes sequences with same gene name) where exists. After we obtain all annotated genomes from dogma, we store it in the local database. The code will then automatically lunch the second step to extract coding genes names and counts. The competition will start by building intersection matrix to intersect genomes vectors in the local database with the others. New core vector for two leaf vectors will generate and a specific \textit{CoreId} will assign to it. an evolutionary tree will take place by using all data generated from step 1 and 2. The tree will also display the amount of genes lost from each intersection iteration. A specific excel file will be generated that store all the data in local database. The whole operation illstrate in Figure \ref{dogma:geneextraction}.
152 \includegraphics[width=0.7\textwidth]{Dogma_geneextraction}
153 \caption{Extract core genes based on Dogma gene names and counts}\label{dogma:geneextraction}
156 The main drawback from the method of extracting core genes based on gene names and counts is that we can not depending only on genes names because of three causes: first, the genome may have not totally named (This can be found in early versions of NCBI genomes), so we will have some lost sequences. Second, we may have two genes sharing the same name, while their sequences are different. Third, we need to annotate all the genomes.
158 \subsubsection{Extracting Core genome from NCBI gene contents}
162 \subsubsection{Core genes based on Dogma Genes names and counts}
