Network Applications

Two examples of IT applications that require their supporting network to be validated will be described here. The first example describes a Manufacturing Execution System (MES) based on a LAN, and the second example desaibes an Enterprise Resource Planning (ERP) application that uses a WAN. 

The system is interfaced with a local LIMS, and data are exchanged hack and forward between the systems. In this example, the GxP nature of the data managed by the MBS and LIMS and data exchange between the systems requires that both systems and the LAN be validated. 

In this example, and generally speaking in all cases in which more than one application shares the same LAN, it is convenient to proceed with a separate network validation project. All of the applications that need to be validated can refer to the validation package of the network, avoiding duplicated work during validation of the IT applications. 

There are about 700 application users, with an average of 300 concurrent users, and 4 MESs exchange data with the SAP R/3. Each manufacturing site installed a different MES based on the topology described in A Site MES Application example above. The ERP application passes Good Manufacturing Practice (GMP)-relevant data (including production orders, bills of materials, materials allocation in the warehouse, materials consumption data, and materials status) back and forward with the MES. 

Due to the client-server nature of the SAP R/3 product, the client side SAP Graphical User Interface (GUI) must be aligned at the server version. To guarantee that all 700 users of the application, distributed on the five sites, receive client version upgrades at the same time and align with the server version concurrently, an application has been installed to automatically distribute the software through the network (Microsoft SMS application). 

---

What is the goal of the network application (classification, modeling, etc.) ... 

Goodacre, R. Edmonds, A. N. Kell, D. B. Quantitative analysis of the pyrolysis-mass spectra of complex mixtures using artificial neural networks Application to amino acids in glycogen. J. Anal. Appl. Pyrolysis 1993, 26, 93-114. 

Goodacre, R. Karim, A. Kaderbhai, M. A. Kell, D. B. Rapid and quantitative analysis of recombinant protein expression using pyrolysis mass spectrometry and artificial neural networks Application to mammalian cytochrome b5 in Escherichia coli. J. Biotechnol. 1994,34,185-193. 

L. Lardon and J.P. Steyer. Using evidence theory for diagnosis of sensors networks application to a wastewater treatment process. In Int. Joint Conf. Artificial Intell. (IJCAI), pages 29-36, Acapulco, Mexico, 2003. 

Continuous random network (applicable to covalent glasses)... 

The client/server model often allows easier integration with other network applications (eg, finance, project management, or human resources) which typically operate in the environment of the server component of the client/server system. Client/server can be gradually introduced in an existing minicomputer environment, often with litde adverse incremental impact in terms of retraining and additional cost. 

G. Hobson, "Neural Network Applications at PSP," paper presented at NPRA Computer Conference, Seattle, Wash., 1990. 

Manallack, D.T. and Livingstone, D.J., Artificial neural networks application and chance effects for QSAR data analysis, Med. Chem. Res., 2, 181-190, 1992. 

Ivanciuc, O., Artificial neural networks applications. Part 7. Estimation of bioconcentration factors in fish using solvatochromic parameters, Revue Roumaine de Chimie, 43, 347-354, 1988. 

From the figure it is easy to see that an infinite number of lines can be drawn that separate the depressed points from the not depressed points in the plane. This is a characteristic of regression and neural network applications rarely is there one solution but a whole family of solutions for a given problem. Nevertheless, a perception can be easily trained to classify patients based on the two hypothetical measures posed. One perception with its trained weights for this particular set of data is shown in Figure 3.2. 

Despite the fact that the neural network literature increasingly contains examples of radial basis function network applications, their use in genome informatics has rarely been -reported—not because the potential for applications is not there, but more likely due to a lag time between development of the technology and applications to a given field. Casidio et al. (1995) used a radial basis function network to optimally predict the free energy contributions due to hydrogen bonds, hydrophobic interactions and the unfolded state, with simple input measures. 

Self-organizing maps are, like most neural network applications, limited by the quality of the data that is used to train them. If the training data is not representative of the whole set of data to which a network is expected to apply, then the clusters found in training may not be representative. This is especially critical with small sets of training data. 

Figure 6.1 Design issues of neural network applications for genome informatics.

Table 9.1 Neural network applications for nucleic acid sequence analysis.

Table 10.1 summarizes neural network applications for protein structure prediction. Protein secondary structure prediction is often used as the first step toward understanding and predicting tertiary structure because secondary structure elements constitute the building blocks of the folding units. An estimated 90% or so of the residues in most proteins are involved in three classes of secondary structures, the a-helices, p-strands or reverse turns. Related to the secondary structure prediction are also the prediction of solvent accessibility, transmembrane helices, and secondary structure content (10.2). Neural networks have also been applied to protein tertiary structure prediction, such as the prediction of the backbones or side-chain packing, and to structural class prediction (10.3). 

Protein secondary structure prediction is one of the earliest neural network applications in molecular biology, and has been extensively reviewed. Typified by the work of Qian and Sejnowski (1988) (Figure 10.1), early studies involved the use of perception or three-... 

Neural network applications for protein sequence analysis are summarized in Table 11.1. Like the DNA coding region recognition problem, signal peptide prediction (11.2) involves both search for content and search for signal tasks. An effective means for protein sequence analysis is reverse database searching to detect functional motifs or sites (11.3) and identify protein families (11.4). Most of the functional motifs are also... 

Integration of Statistical Methods into Neural Network Applications... 

How have neural networks been used in genome informatics applications In Part II, we have summarized them based on the types of applications for DNA sequence analysis, protein structure prediction and protein sequence analysis. Indeed, the development of neural network applications over the years has resulted in many successful and widely used systems. Current state-of-the-art systems include those for gene recognition, secondary structure prediction, protein classification, signal peptide recognition, and peptide design, to name just a few. 

A significant difference between the MHW guidefine and other FDA and EU guidance, however, is that the Japanese guidance only applies to networked applications stand-alone applications such as PECs embedded in equipment and laboratory instrumentation are excused compliance. This position is currently being reviewed, and it is likely that those stand-alone systems that manage electronic records will also require validation. 

Since the start of the new millennium, regulatory authorities have returned to conducting more comprehensive examinations of computer systems. The most significant inspections are typically focused in MRP II or TIMS network applications and include supporting computer network infrastructure. Recent inspections at Eli Lilly, Argus Pharmaceuticals, and Solvay have all taken this approach. The expectations on networks are basic but have often been unsatisfied ... 

The vulnerability of computer services to computer viruses is not easily managed. Besides deploying antivirus software the only other defense is to stop unauthorized software and data being loaded on computer systems and to build firewalls around networked applications. This is a prospective approach that assumes existing computer services are free from computer viruses. However, this approach cannot entirely remove the threat of computer viruses from computer services. The source of authorized software and data may itself be unknowingly infected with a computer virus. Novel viruses can also break through network firewalls. It is therefore prudent to check software and data related to computer services that are used within an organization. 

Staff and management, system support staff, and quality assurance. User-communities of the networked applications should also be represented. 

---