Reviewed by: 
Xiao Lin

Ý 

In 1898, the Italian scientist Camillo Golgi 
found

another type of cellular organelle and named 
"Golgi

Apparatus".

In the 1960's, Palade 
and his colleagues confirmed

Golgi's theory, the Golgi apparatus is a 
cell

structure mainly devoted to processing the 
proteins

synthesized in the endoplasmic reticulum 
(ER).

These typically look 
like a stack of flattened,

smooth, oval membranes, it is thought that 
certain

chemicals that are synthesized in the ribosomes 
move

along the endoplasmic reticulum to the Gol!
gi

apparatus, where they are concentrated and stored

until 
released by secretory vesicles produced by the

outer portion of the 
Golgi apparatus.

The Golgi apparatus 
consists of several flattened

saclike membranes. These sacs sit one on 
top of the

other like a stack of pancakes, the stack 
included

three distinct regions, and each sac in the 
organelle

contains enzymes that modify proteins and present 
in

the Golgi to perform its various synthetic 
activities.

Transforming many 
newly made proteins into mature is

very important role in Golgi 
apparatus.

Some of these proteins 
eventually become membrane

pro!
teins and help with the functions of transport or

self-recognition, 
these proteins are carried on the

outside of the spherical vesicles 
and transported to

the plasma membrane, some these proteins are 
retained

within the cytoplasm for use by the cell.

Other proteins are 
stored inside the vesicles until

they are needed for export. Hormones 
and enzymes are

stored in this fashion until released though 
the

plasma membrane in a process.

The Golgi apparatus 
adds "addresses" to proteins that

are destined to go to 
another organelle, called the

lysosome.

Mitochondria often 
referred to as the powerhouse of

the cell, it supply the cell with 
chemical energy in
the form of the high-energy molecule 
adenosine
triphosphate(ATP).
 
Unlike most other 
organelles, mitochondria contain
DNA, RNA, and ribosomes, and they 
replicate
themselves. It is now thought that mitochondria 
did
not arise as a part of the cell but were 
originally
independently living cells that were taken into 
larger
cells and were able to survive in a jointly 
beneficial
relationship.



Link: http://www.neci.nec.com/~lawrence/papers/face-
tnn97/latex.html
 
Reviewed by: 
Christoforos ChristoforouÝÝÝÝÝÝÝÝÝÝÝÝÝÝ 
 
Face Recognition: A 
Convolutional Neural 
Network
Approach
 
This report discusses 
research on the authors
development of a face recognition system using 
the
convolution Neural Network approach.
 
Convolutional Neural 
network is allows feature
extraction and classification of images 
through a set
of hierarchical set of layers. It takes and 
input
image, it then applies the convolution of theÝ image
(Convolution is a mathematical operation which can 
be
though as the inner product of the image and a window)
 
followed by local sampling of the image. This
sequence of operations 
is repeated resulting to a set
of features that are used by the MLP 
(Multi-Layered
perceptron network) to classify the image. 
Convolution
neural network provides for partial invariance 
to
translation, rotation, scale, and deformation. This
means 
that faces that are translated, rotated, scaled
or deformed slightly 
can be recognized as well. 
 
The system developed 
takes as input the images and
maps then to a Self-Organizing Map for 
Dimensionality
reduction and then Convolutional Neural Network 
is
used for feature extraction and classification. 
Some
variations of these procedu!
res have been tried by
using Karhunen-Lohve Transform for 
Dimensionality
reduction instead of Self-Organizing Maps 
and
Multi-Layered Perceptron for the classification
without 
feature extraction .
 
The system has been used to do some 
sensitivity
analysis in order to determine which parts of 
the
input image are most important for classification
using 
the method of Baluja and Pomerleau. By
introducing some random noise 
to the input images and
measuring the error in system performance were 
able to
identify as regions of higher importance for 
the
classification the eyes, nose, mouth, chin, and 
hair
regions.
 



Link: http://college3.nytimes.com/guests/articles/2001/08/02/860202.xm
l
 
Reviewed by: 
Tawhidul Chowdhury
Ý 
The article reports on how the passengers can be identified 
without 
passport while they are traveling by airplane. The passengers 
no longer have to 
wait in line at passport control and instead they 
will soon be bypassed by 
simply directing their eyes toward a camera, 
a process calledÝ 
"EyeTicket".
 
Passengers initially 
have to enroll and have to scan their iris by 
camera. The image of 
iris is stored into byte code on a server. So, when a 
passenger 
checks in, he has to stand before the camera again. The cameraÝ takes
another image of the passengerís iris and compares 
it with codes stored 
on the serverís database. If it matches, no 
passport need to be shown in 
the airport. The system also tells 
flight number, seat number and other
information need to 
know.
 
The article emphasizes 
the fact that EyeTicket is faster than any 
system, and it can 
identify someone in less than two seconds. EyeTicket gives 
more high 
levels of detail than facial recognition or digital fingerprints. 
For 
example, iris has about 240 unique areas, whereas a face has about 80 
unique areas and fingerprint has about 20 to 40 unique areas. Also, 
Chief 
technology officer, Bill Willis at Iridian technologies in 
Moorestown, New Jersey,
said, EyeTicket is vital for verifying the 
identities of people in
high-security enviro!
nments like airport.
 
Moreover, Eye Ticket 
trial program will begin this fall at Heathrow 
Airport, London, and 
has tested with airline and airport employees in Charlotte,
North 
Carolina.
 



Link: http://www.incyte.com/insidegenomics/foc/foc_foc_0003/foc_foc_00
03_1.shtml
 
Reviewed by: 
Changyou Yu
 
ÝÝÝÝÝÝÝ SNPs Helps Simplify Complex 
Diseases
 
 
SNPs (single 
nucleotide polymorphisms) are the single-letter variations
in genome. 
It is the most common and simple type of genetic variation
shared 
among people and can be used as markers to help geneticists 
associate 
portions of the genome with the incidence of disease.
 
 
Complex diseases such 
as adult-onset diabetes, Alzheimer's disease, and
cancer are caused by 
the combination of variations in genetic sequences
and other 
environmental factors. To discern the genetic roots of 
disease, SNPs 
are used to provide a shortcut for pinpointing certain genes th!
at 
may contribute to complex diseases because many SNPs lie within 
genes
associated with a disease while others are near such 
genes.
 
 
Professor Graeme Bell 
and his group at the University of Chicago 
started to identify the 
genes responsible for adult-onset diabetes in the late
1990s by 
sequencing a 1.7-megabase region ofÝ DNA 
from ten Mexican 
Americans then comparing their DNA to discover SNPs. 
They found a SNP with a 
G(guanine)in place of an A(adenine) at 
position 43. The result turned out that this 
SNP43 does not have an 
obvious connection to diabetes. The group then 
re-sequenced a 66,000-
nucleotide (66-kilobase) stretch in and around SNP43 and found 
more 
polymorphisms. Among them, the combination of three !
SNPs on the same
chromosome (SNP43, SNP19 and SNP63) conferred risk 
for diabetes in 
Mexican Americans. Using similar method, another 
research group based at Duke
University Medical Center led by Dr. 
Margaret Vance and Dr. Jeff Vance
looked for the association of SNP 
variation and the occurrence of
Alzheimer's disease with the DNA 
obtained from 220 patients and 220
age-matched controls. They found 
three SNPs that indicate a direct link 
to the 
disease.
 
 
To successfully use 
SNPs, advances in automation andÝ micro 
arrays can 
be used to provide efficient and cost effective ways to 
screen SNPs in the
population. Genetic analysis will also allow 
researchers to pick out 
more obscure linkage with the aid of mor!
e sophisticated computer algorithms.
 
 
In conclusion, the 
release of the human genome and its accessory SNP 
map provides 
investigators with many more clues to decipher the development 
of 
disease and it would mark an entirely new era in clinical therapy.
 
 



Link: http://www.eurekalert.org/pub_releases/2001-11/nu-
sdm111601.php
 
Reviewed by: Ercan 
Kocabasoglu
 
 
Scientists design 
molecules that mimic nano-structure of bone
 
 
This site describes 
the first design molecules that could lead to a
breakthrough in bone 
repair. The designer molecules hold promise for 
the development of a 
bonelike material to be used for bone fractures or in 
the treatment 
of bone cancer patients and have implications for the 
regeneration of 
other tissues and organs.
 !
The molecules self-assemble into a three-dimensional structure that 
mimics the key features of human bone at the nano-scale level, 
including the
collagen nano-fibers that promote mineralization and the 
mineral
nano-crystals. Collagen is found in most human tissues, 
including the 
heart, eye, blood vessels, skin, cartilage and bone, 
and gives these tissues 
their structural strength. When the synthetic 
nano-fibers form they make a gel 
that could be used as a sort of glue 
in bone fractures or in creating a 
scaffold for other tissues to 
regenerate. Because of its chemical structure, the
nano-fiber gel 
would encourage attachment of natural bone cells, helping 
to patch 
the fracture. The gel also could be used to improve implants or 
hip 
and other joint replacements
 
The findings also map 
out a pat!
h for the creation of many other 
materials by self-assembly and 
spontaneous mineralization that take advantage of an
inorganic 
material growing on an organic material (known as a 
composite) and 
which could be useful in electronics, photonics, magnetics and 
catalysis.
 
The findings also map 
out a path for the creation of many other 
materials by self-assembly 
and spontaneous mineralization that take advantage of an
inorganic 
material growing on an organic material (known as a 
composite) and 
which could be useful in electronics, photonics, magnetics and 
catalysis.
 
To recreate bone's 
nano-structure in the laboratory, Stupp and his team
designed a cone-
shaped molecule, called a peptide-amphiphile, that is
bulkier and 
water-loving on !
one end (a peptide) and slimmer and 
water-phobic on the other (an 
alkyl group). When in water at low pH, the molecules
assemble 
themselves like spokes on a wheel, with the hydrophobic greasy 
tail 
directed to the center, leaving the peptide to face the exterior 
aqueous environment. This basic structure is repeated so that a long 
nano-fiber 
is formed, like an insulated copper wire where the 
insulation is the 
peptide and the wire the alkyl group. The synthetic 
fibers orient the growth of 
the hydroxyapatite crystals so that they 
mimic the structure found in 
natural bone.